Lesson 31
Risk Management 101
Decide what a trade may lose before you buy: a small loss budget, a stop where you are wrong, a share count that follows from both, and a plan for when the stop sells below its price.
Risk management in trading means deciding, before you buy, how much one trade may lose. Set a loss budget, such as 1% of your trading money; find the price that proves you wrong; divide the budget by the distance to that price to get your share count, rounded down. Then remember: a stop decides when you sell, not the price you get.
In November 2024, Swiggy sold its new shares at ₹390 each. On 23 October 2026 the share closed at ₹176.00, its lowest close since listing: 54.9% below what its IPO buyers paid. To get back to ₹390 from there, it must rise 121.6%. A loss and the gain that undoes it are not the same size. That is the first reason to decide your loss before you buy.
By the end you will be able to answer three questions: how much one trade should be allowed to lose, and how many shares that lets you buy; why a stop-loss can sell far below its price, and how to plan for it; and when two holdings are really one risk, and whether buying more of a fallen share is a fix.
Maya, who runs a burger restaurant and is learning to invest, is made up, and so are her amounts: a learning budget of ₹100000 (₹1 lakh) and the rules she writes here. Sun Pharma, Swiggy, IndusInd Bank, Infosys and TCS are real, and every price comes from the daily records of NSE, the National Stock Exchange of India, with its date. Every trade on this page is an illustration on real prices, not a trade anyone made. Amounts are in Indian rupees (₹), and charges and taxes are left out. Where we do our own arithmetic, the figure says "derived".
Start With What Can Go Wrong
Many things can make a trade lose; the size you hold decides what it costs
Maya has written her first research note of her own, on . Before the lunch rush she opens its quote page and asks the question the note left open: how much could this cost me?
Sun Pharma is a huge, steady drug maker. If I buy some, what could really go wrong?
Look at its own year. From 30 July to 19 October 2026 it fell 12.9%, with no crisis.
12.9%! Then how much could I lose?
That depends less on Sun Pharma than on you. How many shares will you hold?
Like a new burger. One tray to test it, not the whole week's order.
Exactly. Decide the loss first. The size follows from it.
Figure 1The risk dashboard
Decide the loss first: ₹1000 a trade
Maya's learning moneymade up
1%of the learning budget, the most one trade may lose
What can go wrong. Many things can make a trade lose money, and none of them needs the company to fail. The price can fall on the company's news or the whole market's (what moves share prices). It can jump past your exit overnight (chapter 5). Nobody may buy when a share is locked at its limit (circuit limits and surveillance). Several holdings can fall together (chapter 7). You can add more to a loser (chapter 8). Borrowed money makes any fall bigger (margin, leverage and pledging). And a typing slip or a fraud can cost you directly (placing your first order and staying safe).
The floor and the ceiling. With shares paid for in full, the most you can lose is what you paid. With borrowed money, or a short sale (selling shares first and buying them back later), you can lose more. Risk management cannot stop a price falling; it decides, before you buy, how much a fall may cost you.
Size sets the cost. The rupees lost are the shares held times the fall per share. Sun Pharma had its of the year on Thursday 30 July 2026, ₹2001.10. On Monday 19 October it closed at ₹1743.00: ₹258.10 lower, a fall of 12.9%, with no crisis. Say someone had bought at that high close and held. On 5 shares the fall cost ₹1290.50; on 25 shares, ₹6452.50; on 49 shares, which cost nearly all of a ₹1 lakh budget, ₹12646.90 (derived). Same company, same days, same fall: only the size changed.
Figure 2Size decides the loss
One fall, three sizes
Thu 30 Jul 2026₹2001.10Mon 19 Oct 2026₹1743.00−12.9%₹258.10 a share, the same on every row1.29% of a ₹100000 budget (derived)
6.45% of a ₹100000 budget (derived)
12.65% of a ₹100000 budget (derived)
And if it falls a lot, it just climbs back the same way, doesn't it?
Recovering a Loss Needs a Bigger Gain
A fall is measured on what you had; the climb back on what is left
Swiggy's IPO buyers paid ₹390, and it closed at ₹176.00. A 54.9% rise gets them back, right?
No. From ₹176.00 it needs +121.6% to reach ₹390 again.
More than double the fall? The rupees are the same both ways.
The rupees, yes. The base is not: the climb starts from what is left.
One real road. priced its IPO shares at ₹390 in November 2024; one lot was 38 shares, ₹14820. On Friday 23 October 2026, the day after its July to September results, it closed at ₹176.00, its . That lot was now worth ₹6688: ₹8132 less, a fall of 54.9% (derived).
Why the climb is bigger. To get back to ₹14820, the ₹6688 that is left must earn ₹8132, and ₹8132 is 121.6% of ₹6688. The rupees are the same both ways; the base is not.
The rule. Gain needed = the fall ÷ what is left. Down 10%, up 11.1% to get back; down 25%, up 33.3%; down 50%, up 100%; down 75%, up 300%. A small fall needs about the same gain back; a deep one needs far more, and the gap widens fast.
Figure 3The climb back
Down 54.9%, up 121.6% to get back
One IPO lot, 38 shares: ₹390 → ₹176.00
Figure 4The gain needed
Down 10%, up 11.1%. Down 75%, up 300%.
gain needed to get back, for each fallif the climb were the same size
What the arithmetic does not say. It says nothing about whether a recovery comes, how long it takes or what Swiggy does next. A price owes its old level nothing (why the stock market exists).
So the trick is never to fall that far. How small should one loss be?
A Loss Budget per Trade
A small, fixed share of the money set aside for trading, decided before each trade
₹1000 a trade sounds tiny. Wouldn't 10% of my ₹1 lakh teach me faster?
Ten 10% losses in a row would leave you ₹34868.
And ten at 1%?
₹90438. Small enough to keep going, and to keep thinking.
Where the money comes from. In asset allocation basics Maya set aside a learning budget of up to ₹100000: money she could lose without touching her emergency fund, her goals or the restaurant. Only this money is ever at risk in a trade.
Her budget. Maya chooses 1%: no trade may lose more than ₹1000. Many trading guides use 1% to 2%; it is a , not a rule, and smaller is allowed.
Not what she invests. ₹1000 is the loss she accepts if she is wrong. The shares she buys can cost far more, as chapter 4 shows. The budget caps the damage, not the purchase.
Why so small: a run of losses. Losses come in runs, for reasons nobody controls. After ten losses in a row, each 1% of what is left, ₹100000 becomes ₹90438, and +10.6% gets it back. At 10% a loss it becomes ₹34868, and the climb back, by chapter 2's rule, is +186.8% (derived). That is what a run does, not how likely one is.
Figure 5A loss budget
Ten losses in a row: ₹90438 left, or ₹34868
at 1%: ₹99000
at 10%: ₹90000After 2 losses in a row
at 1%: ₹98010
at 10%: ₹81000After 3 losses in a row
at 1%: ₹97030
at 10%: ₹72900After 4 losses in a row
at 1%: ₹96060
at 10%: ₹65610After 5 losses in a row
at 1%: ₹95099
at 10%: ₹59049After 6 losses in a row
at 1%: ₹94148
at 10%: ₹53144After 7 losses in a row
at 1%: ₹93207
at 10%: ₹47830After 8 losses in a row
at 1%: ₹92274
at 10%: ₹43047After 9 losses in a row
at 1%: ₹91352
at 10%: ₹38742After 10 losses in a row
at 1%: ₹90438
at 10%: ₹34868
losses in a row
Recompute after losses. 1% is 1% of what is left: at ₹95000, the budget is ₹950. Keeping the old ₹1000 quietly raises the risk.
₹1000 it is. So how many Sun Pharma shares does that let me buy?
From Budget to Share Count
Decide where you are wrong, divide the budget by the distance, round down
I'd planned to buy about ₹20000 of Sun Pharma and put a stop somewhere below.
That's backwards. Find where you're wrong first; the share count comes last.
Say I'm wrong if it falls 5%: that's ₹1655.85, ₹87.15 a share below today's close.
₹1000 ÷ ₹87.15 is 11.47. Round down: 11 shares.
Where am I wrong? The exit price comes first, and it comes from your reason for buying: for a trade, a level on the chart that would prove the reason wrong (charts and price action). A research note's lines are different: they say when to read again, not where to sell (how to research a stock). For this illustration Maya writes a simple rule: if Sun Pharma falls about 5% from where she buys, her reason was wrong. On the close of Monday 19 October 2026, ₹1743.00, that is ₹1655.85. That price is her stop: if the share trades there, a stop-loss order sends her sell order.
The loss per share: ₹1743.00 − ₹1655.85 = ₹87.15.
Divide, and : ₹1000 ÷ ₹87.15 = 11.47, so 11 shares. Never 12: rounding up breaks the budget you just set. The planned loss is 11 × ₹87.15 = ₹958.65.
Check the money: 11 × ₹1743.00 = ₹19173.00, about a fifth of her learning budget. Charges come on top (brokerage charges and taxes); a careful plan leaves a little room for them.
The stop decides the size. The same ₹1000 with a stop just under the past week's low (₹1725.50 on 14 October), at ₹1725.00, is only ₹18.00 a share: 55 shares, ₹95865.00, nearly all her learning budget in one share. That level was reached in the very next session: on Wednesday 21 October (Tuesday was a holiday) the lowest trade was ₹1721.10, and the day closed at ₹1745.10. With a stop just under the year's low (₹1583.70 on 29 January 2026), at ₹1583.00, it is ₹160.00 a share: 6 shares, ₹10458.00 (derived). A closer stop is not safer: an reaches it, and it asks for more money.
Figure 6From budget to share count
The same ₹1000, three stops, three sizes
₹1743.00buy at the close, 19 Oct₹1743.00 − ₹1655.85 = ₹87.15 a share₹1000 ÷ ₹87.15 = 11.47, rounded down: 11 shares11 × ₹1743.00 = ₹19173.00planned loss ₹958.65
Two limits, and the smaller wins. The share count must pass both the loss budget and the cash you have. The arithmetic never creates money: if the count costs more than your cash, buy fewer, and never borrow to fit it (margin, leverage and pledging). To try your own numbers, use the position size calculator.
So if Sun Pharma drops to ₹1655.85, my stop sells there and I lose ₹958.65?
A Stop Is Not a Promised Price
A stop decides when you sell; the next buyer decides the price
Circuit limits and surveillance showed that on a share locked at its lower limit, a stop sends an order, not a buyer. This chapter is about an ordinary day, with buyers all day long. Even so, the price was the problem: the stop fired, but nobody bought near it.
The plan, an illustration on real prices. Say someone sized a trade exactly as Maya did: , a private bank, at the close of Monday 10 March 2025, ₹900.50; a stop about 5% lower, at ₹855.50; ₹1000 ÷ ₹45.00 = 22 shares, ₹19811.00 of stock. Planned loss: ₹990.00.
That evening, after the close, at 6:36 pm by NSE's record, the bank told the exchanges it had found discrepancies in some account balances of its derivatives portfolio, an estimated hit of about 2.35% of its net worth.
The next morning, the first trade on Tuesday 11 March was at ₹810.45, 10% below Monday's close, and the share never traded higher that day. It closed at ₹655.95, 27.2% lower. A stop that sells at the market could sell no higher than ₹810.45: a loss of at least ₹1981.10, twice the plan. A stop with a limit at ₹850 never sold, because every trade was below its limit; held to the close, the loss on paper was ₹5380.10, more than five times the plan (derived). Whether the stop was a day order placed again when trading opened at 9:15 am, or a standing the broker kept, it could not sell at ₹855.50: nobody traded there. Placing your first order has the order types.
Figure 7A stop is not a promised price
No trade between ₹900.50 and ₹810.45, where the stop was
IndusInd Bank: 22 shares at ₹900.50, stop ₹855.50, the same ₹1000 plan illustration
What a stop can and cannot do. It acts when you are not watching; that is its use. It cannot pick the price, and in a fast market the fill "could be markedly different" from the stop price, as FINRA, the US brokers' regulator, puts it. News arrives overnight, on results days and at weekends.
Plan for it. Keep the budget small enough that two or three times the planned loss would still hurt, not harm. Know when the company reports. And never size so large that a gap could take money you need.
Figure 8A stop is not a promised price
Planned ₹990.00. Lost at least ₹1981.10.
IndusInd Bank: 22 shares at ₹900.50, stop ₹855.50 illustration
A stop decides when you sell. The next buyer decides the price.
If losses can run past the plan, how do I know a plan is worth repeating at all?
Reward-to-Risk and Expectancy
A ratio says how often a plan must work; only a record says whether it does
A 3 to 1 trade sounds like a 75% chance of winning.
It's a plan, not a chance. It says how often it must work just to break even.
And how often is that?
One trade in four. Whether yours manage it, only your record can say.
R. Maya's planned loss on any trade, about ₹1000, is her . A trade that gains ₹2000 made 2R; one that hit its stop made −1R; the gap in chapter 5 cost about −2R.
Reward-to-risk compares the planned gain with the planned loss. It is a plan, not a forecast and not a chance.
The break-even line. If every win were 1R and every loss 1R, half the trades must win just to break even. At 2R wins, one in three (33.3%); at 3R, one in four (25%). A bigger target needs fewer wins, and it is usually reached less often; the ratio says nothing about how often.
Costs move the line. If each trade costs 0.1R in charges, a 2R win nets 1.9R and a loss costs 1.1R: break-even rises from 33.3% to 36.7% of trades (derived).
Figure 9Reward-to-risk
How often a plan must work, just to break even
fewer wins than this: the plan loses moneymore: it pays its way
Expectancy is the average result per trade: the share of wins times the average win, minus the share of losses times the average loss, after costs. Say a made-up record of ten trades at 2 to 1 shows four wins of 2R and six losses of 1R: (8R − 6R) ÷ 10 = +0.2R a trade, about ₹200 on Maya's ₹1000, before costs (derived). Real expectancy comes from a record of real trades, written down (the journal in build your first watchlist), never from targets. A short record says little; a few lucky trades can look like skill (trading psychology).
Then I'll spread out: Infosys and TCS, ₹1000 each. Two names must be safer, right?
Several Names, One Risk
Count causes, not names
Infosys and TCS are two different companies. Two trades, two budgets: that's spreading my risk.
What does each one depend on, Maya?
Both sell IT work to big clients, many in America, with engineers from India. Oh. The same thing.
So one rule there can reach both on the same morning. It did, in September 2025.
On Friday 19 September 2025 (US time), the US president signed a proclamation requiring a $100000 payment with new petitions for H-1B visas, the permits many companies use to bring skilled workers to the US, from 21 September, as law firms summarised it.
On Monday 22 September, the first Indian session after it, closed at ₹1499.50, 2.64% lower, and TCS at ₹3073.80, 3.01% lower. The Nifty IT index fell 2.95%; the Nifty 50, India's main index, fell 0.49%. Reports that day tied the IT fall to the visa fee. Two ₹1000 budgets on the two names were one ₹2000 bet on one story.
Figure 10Several names, one risk
One rule, two names, one morning
Infosysclosed ₹1499.50−2.64%
TCSclosed ₹3073.80−3.01%, in plain words, is how much two prices tend to move together. The useful question is simpler: do my holdings depend on the same customers, rule, currency, raw material or mood?
Group by cause, and give each cause one budget, not each name. Two IT companies, two banks or two cement makers are often one bet.
What spreading can and cannot do. Owning businesses with different drivers softens one company's or one industry's bad day; it cannot remove a fall in the whole market. A broad fund spreads further than a handful of names (mutual funds, index funds and ETFs); the plan for it is in asset allocation basics.
And if one of them falls, can't I buy more at the lower price to bring my average down?
Averaging Down
The average falls; the money on the same idea grows
If that IPO buyer buys 38 more Swiggy shares at ₹176.00, the average drops to ₹283.00.
And the money in rises to ₹21508. Each ₹10 fall now costs ₹760, not ₹380.
So the climb back looks shorter, but the bet is twice as big.
Ask one thing: would you buy those shares today, if you owned none?
If a new dish isn't selling, cooking twice as much of it doesn't fix the menu.
The tempting move, an illustration on real prices. Take chapter 2's IPO buyer: 38 Swiggy shares bought at ₹390, ₹14820, worth ₹6688 at ₹176.00. Say they buy another 38 at ₹176.00, for ₹6688. Now 76 shares have cost ₹21508: an of ₹283.00 a share.
What looks better. To get back to the average, the price now needs +60.8%, not +121.6%.
What actually changed. The ₹8132 already lost is still lost. The money riding on Swiggy's next move doubled, and so did the cost of every further fall: each ₹10 now costs ₹760 instead of ₹380 (derived). Whether the new shares help depends on what the price does next, which nobody knows.
Figure 11Averaging down
The average fell. The bet doubled.
Price now ₹176.00, Fri 23 Oct 2026 illustration
The test. Ask: if I owned none, would I buy these shares today, with this stop and this budget? If yes, it is a new trade: size it as in chapter 4, and count it with the first under one cause, as in chapter 7. If the honest reason is the average, it is not a plan.
Not the same as a planned purchase. Buying a broad fund every month on a schedule written in advance (mutual funds, index funds and ETFs) is a plan made before any fall; adding to one share because it fell is a decision made inside the loss (trading psychology explains why it feels right).
Each trade has its budget now. But what stops me when everything goes wrong at once?
A Pause Rule and the Whole Account
Add up everything open, and write the line where you stop
Five trades at ₹1000 each. That's ₹5000 that could go at once.
If every stop fills at its price. Now write the line where you pause.
₹90000. Below it, no new trades until I've read my journal.
Write it now, while you're calm. Not during the losses.
. Add the planned losses of everything open. Say Maya has five trades open, each with a ₹1000 planned loss: ₹5000, 5% of her learning budget. If two of them share a cause, chapter 7 says to count them as one bet, and chapter 5 says any of them can lose more than planned.
Maya's pause line, written before she needs it. If her learning money falls 10% below its highest point, from ₹100000 to ₹90000, she opens no new trade until she has reviewed her journal: which losses followed her rules, which broke them, which fills were worse than planned.
Figure 12The whole account
Everything open, if every stop is hit, stays above the line
10% below the peak: no new trades; review the journal firstroom left ₹5000
planned loss ₹1000
same cause as trade 4Trade 4
planned loss ₹1000
same cause as trade 5Trade 3
planned loss ₹1000Trade 2
planned loss ₹1000Trade 1
planned loss ₹1000
made up Maya's rule, written before she needs it: "If my learning money falls 10% below its highest point, to ₹90000, I open no new trade until I have reviewed my journal."
A pause stops new trades. It does not sell what is open, and it does not cancel stops.
What a pause is. It stops new trades; it does not sell what is open, and it does not cancel stops. It ends with a reason to restart, never with a wish to win the money back (trading psychology names that urge).
Her numbers are hers. 1%, five trades and 10% are Maya's choices for her money; yours should fit your money and your nerves. What matters is that they are written down first.
One thing could break all of it: money that is not hers. Maya once turned down a bank loan for her restaurant. Borrowing to trade is a loan too, and it makes every fall in this lesson bigger. Is it the same loan she refused? That is the question of margin, leverage and pledging.
Before you go
Check yourself
4 questions from this lesson. Pick an answer to see why, then slide on to the next.
Question 1 of 4
Questions readers ask
Common questions
What is risk management in trading?
Deciding, before you buy, how much one trade may lose, and sizing the trade to fit. You set a loss budget, find the price that proves you wrong, divide to get the share count, plan for a stop filling worse than its price, and add up the risk of everything you hold.
How much should I risk per trade?
There is no rule. Many trading guides use 1% to 2% of the money set aside for trading; on ₹100000, 1% is ₹1000. Smaller leaves room for runs of losses: ten 1% losses in a row leave ₹90438, ten 10% losses leave ₹34868.
How do I calculate position size?
Shares = loss budget ÷ (buy price − stop price), rounded down. With ₹1000, and Sun Pharma at ₹1743.00 with a stop at ₹1655.85 (19 October 2026), that is ₹1000 ÷ ₹87.15 = 11 shares, ₹19173.00 of stock. Then check that you have the cash.
Does a stop-loss guarantee my exit price?
No. It sends a sell order once a trade reaches the trigger; the price is whatever the next buyer pays. On 11 March 2025 IndusInd Bank first traded at ₹810.45 after a ₹900.50 close, so a stop at ₹855.50 could not sell anywhere near its price.
Why does a 50% loss need a 100% gain to recover?
Because the gain is measured on what is left. ₹100 that falls 50% is ₹50, and ₹50 must double to be ₹100 again. Swiggy, down 54.9% from its ₹390 issue price on 23 October 2026, needs +121.6%.
Is averaging down a good idea?
It lowers your average price but adds money to the same idea and doubles what each further fall costs. Ask whether you would buy those shares today if you owned none; if yes, size it as a new trade.
Words to keep
The 18 words this lesson used, in plain English
Risk
In this lesson, the rupees a trade could lose, decided before you buy. Sun Pharma's 12.9% fall cost ₹1290.50 on 5 shares and ₹12646.90 on 49.