What Is Risk Management in Trading? 1% vs 5% Risk on 33 Years of SPY

Michael Whitman · Trading since 2017, full-time since 2020 · YouTube creator, 137K subscribers · Published 2026-09-05 · Updated 2026-11-02 · Reviewed 2026-11-01

Risk management in trading is the set of decisions that fix what a trade can cost you before you place it: where you get out if you are wrong, and how big the position is so that exit costs a set share of your account. It does not pick winners. It decides how much damage the losers can do, and on real data that difference is large.

Most guides stop at “risk 1% per trade” and a made-up example. This page runs the same 1,758 real SPY trades through four risk levels, from 0.5% to 5% of the account, so you can see what the choice does to the worst stretch of losses rather than take it on faith.

How it works

There are three choices, and the order matters more than any of them.

First, the price where the trade idea stops being true. It has to be a level you can point to on the chart, such as the low that the move started from, not a round number of dollars you would be comfortable losing. That level becomes your stop loss. Be wary of the spot just under an obvious low: stops tend to cluster there, and the liquidity page covers why that matters.

Second, the share of the account you accept losing at that price. This is a standing number, picked once while no position is open: 0.5%, 1%, 2%. It is the same on the trade you love and the trade you are unsure about.

Third, the position size, which is not a choice at all. Divide the money you accept losing by the distance from entry to stop, and the number of shares is whatever comes out. The position sizing page goes through that division step by step.

1R is the name for the distance from entry to stop. Measuring results in R (a loss of 1R, a gain of 0.5R) lets you compare trades on a $20 stock and a $700 fund on the same scale, which is how the test further down is scored.

Where the planned loss and the real loss part ways

The arithmetic assumes the stop fills at the stop. Usually it does. When a market opens below your stop, the order fills at the first price available, and that can be well beyond the line.

In the 1,758-trade test, 290 trades were stopped out at their stop price, 35 opened beyond it, and the rest were still open at the end of the week and closed there. Those 35 gap exits are the only trades that lost more than 1R. The worst, in the week of 11 January 1999, lost 1.75R: a planned 1% loss would have come out as about 1.75%.

That is why risk per trade is a planning number with a margin of error, and why the margin grows with holding time. A position held over a weekend or an earnings report carries a gap the stop cannot catch.

A worked example: one SPY trade in December 2019

This is a real trade from the test, on a $10,000 account risking 1%, so $100.

The setup. SPY opened Monday 2 December 2019 at $314.59. The average daily range over the prior 14 sessions (the ATR) was $1.73, so the stop sat two ranges lower, at $311.12. The distance from entry to stop was $3.47.

The size. $100 divided by $3.47 is 28.8 shares. Rounding down to whole shares gives 28, which puts $97.12 at risk if the stop fills where it was placed. Those 28 shares cost $8,808.52, about 88% of the account. A 1% risk does not mean a 1% position. Here it meant most of the money was in the trade, because the stop was close.

What happened. Monday’s low was $311.17, five cents above the stop. On Tuesday 3 December SPY opened at $308.65, below the stop, so the stop filled at the open. The loss was $166.32, or 1.71R, about 1.66% of the account instead of the planned 0.97%.

The same trade at 5% risk. $500 at risk over a $3.47 distance buys 144 shares, a position worth $45,300.96 on a $10,000 account. That needs about $35,000 of borrowed money, more than the 50% down payment of Regulation T allows in a standard US margin account. The Tuesday gap would have cost $855.36. Nothing about the chart was different; only the risk number changed.

These prices check out on a second source: Nasdaq’s historical data shows the same open, high, low and close for SPY on 29 November, 2 December and 3 December 2019.

Why the risk number matters more than the setup

A setup decides whether the average trade gains or loses. The risk number decides how far the account falls while that average plays out, and the two are separate questions. You can have the first one right and still lose the account to the second.

The reason is compounding on the way down. Each loss is a share of what is left, so a run of losses removes more in total than the same run of gains adds back. Losing 10 trades in a row at 1% leaves about 90% of the account. At 5%, the same run leaves about 60%. The risk of ruin page walks through streaks in the abstract; the data below uses the streaks SPY actually produced.

The longest losing run in the test was 10 trades, from the week of 18 August 2008 to 24 October 2008, adding up to a loss of 6.16R. At 1% risk that run cost 6.0% of the account. At 5% it cost 27.0%. Same trades, same dates, same stops.

The original data

The test, exactly as coded. One trade per calendar week on SPY from February 1993 to October 2026, 1,758 trades in all. Buy at the open of the week’s first session. Put the stop two 14-day ATRs below the entry, where the ATR is the simple average true range of the 14 sessions before the entry day. If a later session opens at or below the stop, exit at that open; if a session’s low touches the stop, exit at the stop; otherwise exit at the close of the week’s last session. Prices only: no dividends, no commissions, no interest on borrowed money, fractional shares allowed. This is not a strategy we suggest; it is a simple way to get a long, mixed run of real wins and losses with a fixed stop rule.

What the trades did. 937 gained, 816 lost and 5 closed flat. The average trade gained 0.047R and the median gained 0.078R, which is a small edge and mostly reflects SPY rising over 33 years. Each account starts at $10,000 and risks a fixed share of whatever it holds at the time.

The worst drawdown was the same stretch at every risk level: from a peak after the trade that closed on 17 July 1998 to a low on 7 March 2003. Only the depth changed.

Read the red line as time spent in a hole. The 5% account was more than 20% below its high after 53.7% of all trades, and more than 10% below after 72.6%. The 1% account was more than 20% below after 4.3% of trades and more than 10% below after 24.3%.

The 5% account ended about nine times larger, and that has to be said plainly. The trades had a small positive average, and a larger fraction of a positive average compounds faster. It also needed borrowed money: the median 5% position was 2.14 times the account, and the position was larger than the account in 92.8% of weeks. At 1%, the median position was 0.43 times the account and went over it in 0.85% of weeks.

The test also lends that money for free. It charges no interest and ignores Regulation T’s 50% down payment, which caps a standard margin account at twice its equity; the 5% position broke that cap in 54.1% of weeks. At an assumed 5% a year on the borrowed part, the 5% account would have ended near $48,000; at 8%, near $21,000, below the 1% account. Those numbers describe one sequence that happened. They are not a forecast, and a sequence with a slightly negative average would have punished the 5% account far harder.

One history is one draw, so we reshuffled it. We drew 250 trades at random, with replacement, from the 1,758 real results, 10,000 times for each risk level, and recorded the worst drawdown of each run. 250 weekly trades is roughly five years.

The median worst drawdown across those runs was 4.45% at 0.5% risk, 8.72% at 1%, 17.09% at 2% and 38.86% at 5%. Going from 1% to 2% roughly doubles the typical worst fall, and going to 5% takes a 50% drop from something that did not occur in 10,000 tries to something that occurred in about one run in five. The full trade list, with both account balances after every trade, is in the research file.

Three ways a good plan gets undone

The stop moves away from price

Tightening a stop as a trade moves your way is management. Widening it while the trade moves against you cancels the first decision on the list, and with it the size that was built on it. The loss is no longer the one you planned; it is open-ended, and it feels like patience while it happens.

The stop is jumped to breakeven at the first green tick

Pulling the stop up to the entry price the moment a trade shows a small profit swaps a level the chart gave you for a level only your own entry knows about. Price has no reason to respect it, and trades get closed for nothing that would have reached their target.

Size follows confidence

The trade you feel surest about is the one most people size up. That puts the largest losses on the trades with the strongest conviction, which is backwards from any plan. In the method above, size comes from the stop distance and the risk number only, and how you feel about the trade is not an input.

Questions people ask

What is the 1% rule in trading?

It caps the planned loss on any one trade at 1% of the account, so $100 on $10,000 if the stop fills at its price. It limits the loss, not the position: in the December 2019 example, 1% risk put about 88% of the account into the trade. How the percentage gets chosen is covered on the risk per trade page.

Is 2% risk per trade too much?

It depends on what drawdown you can live through, and the data puts a number on it. On the same SPY trades, 2% produced a worst drawdown of 45.0% against 25.4% at 1%, and a third of the five-year reshuffles fell 20% or more at 2%, against 1.33% at 1%.

What are the main parts of risk management in trading?

A stop at the price where the idea is wrong, a fixed share of the account at risk on each trade, a position size worked out from those two, and limits that sit above single trades, such as a cap on how much you lose in a day or a week. How much money you start with sets what those percentages buy; the trading capital page covers that.

Does a stop loss guarantee my maximum loss?

No. A stop becomes a market order once its price is reached, and if the market opens beyond it you get the opening price. In the test, 35 of 1,758 trades did that, with losses up to 1.75 times the planned amount.

When it fails

Gaps and fast markets

Everything above prices the stop at the stop. Weekends, earnings reports, sudden news and slippage in a thin market can all fill it worse, so a plan that risks 1% should expect the occasional loss of 1.5% or more.

Positions that move together

Five open trades at 1% each are not five separate 1% risks if they all fall on the same day. Stocks in one sector, or a stock and an index fund that holds it, can stop out together. The real risk is the combined loss if everything that tends to move together does, which is where correlation comes in.

One history, one market

The test is one market, one stop rule and one 33-year period in which SPY rose a lot. A strategy with a different average, longer losing runs or wider stops would give different numbers. The point that survives is the shape: the same trades produce a far deeper hole at a higher risk number.

Rules written down and then not followed

Many failures in practice are decisions made with a position open: a stop widened, a size doubled to win back a loss. Writing the stop and the size down before the order goes in is the plainest defense against that, and it means nothing unless the written number is the one you use.

Related

Support and resistance is where most stop prices come from.

Market structure gives you the swing low or high whose break ends the trade idea.

Candlesticks help you tell a level broken on a close from one only poked by a wick.

The drawdown page covers how long falls like the ones above can last.

This page is educational, not financial advice. Test every idea on your own charts before risking money.