How much to put in one trade
Beginner · 2 min
Almost everyone picks the share first and the amount second, usually by feel. Doing it the other way round is the single change most likely to keep a beginner in the game long enough to learn anything.
You cannot control whether a trade works. You can control exactly how much it costs you when it does not. That is what position sizing is: working backwards from the loss you are willing to take to the number of shares you should buy.
Work backwards, not forwards
- Decide what your whole trading capital is. Be honest — money you have actually set aside for this, not your savings.
- Decide what percentage of it you are willing to lose on any one trade. Many people use one or two percent.
- Decide the price at which you would admit the trade is wrong and get out.
- Divide the rupees you are willing to lose by the loss per share. That is how many shares to buy.
Sizing a trade on ₹2,00,000 of capital
- Trading capital
- ₹2,00,000
- Willing to lose on one trade
- 1% = ₹2,000
- Price you would buy at
- ₹450
- Price at which you would get out
- ₹430
- Loss per share if you are wrong
- ₹20
- Shares to buy
- ₹2,000 ÷ ₹20 = 100
- Money actually committed
- ₹45,000
Notice that the ₹45,000 committed is not the risk. The risk is ₹2,000, because you have already decided where you leave.
Try this in the app
- Position size calculator — do this sum without the arithmetic
Share this lesson
Works with
- Support and resistance — Prices where a share has repeatedly stopped falling, or stopped rising.
- The rules around the key point — What you buy, how much, and when you add are separate decisions the key point does not make for you.
- Buying after it goes quiet — Three rules people use to act on ground volume, and why none of them tells you the bottom.
- Trading a hammer — Enter on the confirming close, put the stop under the hammer's low, and size the trade so that stop costs what you planned.
- Where to put the stop — Put it where the reason you bought stops being true.
- When a tip is dressed up as analysis — A pitch that shows you arithmetic and charts can still be a sales pitch, and the arithmetic proves nothing.
- The formula that only works backwards — A projection drawn from two past prices is a ruler, not a forecast.