profit factor meaning in a backtest

Profit factor is gross profit divided by gross loss across every trade in the test, so a profit factor of 1.6 means the winners brought in 1.6 times what the losers gave away. Anything above 1 is profitable before costs are deducted. What the number hides is concentration — one enormous winner can carry it, so check whether the figure survives deleting the best few trades.

annual return divided by maximum drawdown is known as

The MAR ratio. It sets what a strategy earns in a year against the worst peak-to-trough fall it put you through, so 22% a year with an 18% maximum drawdown gives roughly 1.2, while the same 22% with a 52% drawdown gives about 0.4. It is one of the more honest single numbers on a strategy report, because unlike total return or win rate it is hard to flatter.

is a high sharpe ratio over few trades meaningful

Enough that the result cannot be explained by luck — a 2.6 Sharpe over 34 trades sits comfortably inside what randomness produces, while the same figure across several hundred trades is a genuine claim. Sample size is the precondition for every other metric mattering, and it is the number most often left out of a report. A useful companion test is to delete the best 5% of trades and rerun: a real edge degrades gracefully, a curve-fitted one collapses.

what is a good sharpe ratio for a trading strategy

There is no official threshold — above 1 is commonly described as good and above 2 as excellent, but the figure is meaningless without the trade count behind it. Sharpe measures return per unit of volatility and penalises upside volatility exactly as heavily as downside, which quietly flatters strategies that sell tail risk. Read it beside maximum drawdown and sample size rather than on its own.

why do Indian equity backtests starting in 2003 look so good

Because 2003 and 2013 both sit near the bottom of long bull runs, so almost any long-biased strategy tested from those dates looks excellent. Ask to see 2008, 2011, 2018 and 2020 broken out separately — a report showing only the aggregate is hiding the periods that decide whether the system is survivable. Live drawdowns also tend to run worse than backtested ones, because slippage widens precisely when markets move fastest.

Sharpe ratio, profit factor and drawdown

Every strategy comes with a page of statistics designed to look impressive. Most of them can be improved by making the strategy worse, which is why knowing what each one actually measures is the difference between evaluating a system and admiring one.

The metrics, and what each hides

Metric	What it measures	What it hides
Total return	How much it made overall	Everything — the path, the risk, the leverage
Win rate	How often it was right	Payoff size. A 90% win rate with one catastrophic loss is a disaster
Profit factor	Gross profit ÷ gross loss	Whether one enormous winner carried it
Sharpe ratio	Return per unit of volatility	Penalises upside volatility equally; flatters strategies that sell tail risk
Maximum drawdown	The worst peak-to-trough fall	How long it lasted, and how many times it nearly happened
MAR ratio	Annual return ÷ max drawdown	Little — this is one of the more honest single numbers

The number most reports omit

Sample size. A strategy with a 2.4 Sharpe over 31 trades tells you almost nothing — that result is comfortably achievable by chance. The same Sharpe over 800 trades is a genuine claim.

Worked example

Two identical-looking reports

Both show 34% annual return, 1.9 Sharpe

Strategy AFour trades produced 80% of the profit27 trades over 6 years

Strategy BProfit distributed across many trades740 trades over 6 years

What A really isRemove the best three and it is negativeA few lucky trades with a curve fitted around them

What B really isRemoving the best three barely changes itA repeatable edge

The testIf the edge vanishes, it was never an edgeDelete the top 5% of trades and rerun

That last line is the single most useful diagnostic on this page. A real edge degrades gracefully when you remove its best trades; a fitted one collapses.

Loading interactive demo…

Run the same win rate and payoff repeatedly and watch how different the outcomes look over small samples. That variation is exactly what a 30-trade report is showing you.

Drawdown is the number you actually live through

Return is what a report advertises; drawdown is what you experience. A strategy returning 40% a year with a 55% maximum drawdown is, for almost every human being, untradeable — because the drawdown arrives first and you stop.

Two strategies, same annual return

Strategy A — 22% a year

Maximum drawdown 18%
Longest flat period 9 months
MAR ratio 1.2
Genuinely followable

Strategy B — 22% a year

Maximum drawdown 52%
Longest flat period 26 months
MAR ratio 0.4
Abandoned by most people before the recovery

Check yourself

A strategy shows a 2.6 Sharpe ratio over 34 trades. What is the correct conclusion?

Simple bhasha mein

Kitne trade pe yeh result hai?

Koi report dikhaye — 34% return, shaandaar ratio. Pehla sawaal: kitne trade pe? 30 trade ka matlab kuch nahi, woh kismat se bhi ho jaata hai. Doosra: sabse bade 5% trade hata do aur dobara chalao. Asli edge thoda kam hota hai; nakli edge poora gaayab ho jaata hai.

What to remember

Most strategy metrics can be improved by making the strategy worse.
If you read only two numbers, read maximum drawdown and number of trades.
Delete the best 5% of trades and rerun — a real edge degrades gracefully.
Drawdown, not return, is what you actually live through.
Indian backtests starting in 2003 or 2013 begin near a bottom and flatter everything.

You reached the endMark it done and keep your streak going.

Up nextWhat a failed pattern tells you Previous: Where to take profit

Finished this lesson?

Mark it done to track your progress through the curriculum.

Common questions

Short, direct answers to what people ask about this topic.

profit factor meaning in a backtest: Profit factor is gross profit divided by gross loss across every trade in the test, so a profit factor of 1.6 means the winners brought in 1.6 times what the losers gave away. Anything above 1 is profitable before costs are deducted. What the number hides is concentration — one enormous winner can carry it, so check whether the figure survives deleting the best few trades.
annual return divided by maximum drawdown is known as: The MAR ratio. It sets what a strategy earns in a year against the worst peak-to-trough fall it put you through, so 22% a year with an 18% maximum drawdown gives roughly 1.2, while the same 22% with a 52% drawdown gives about 0.4. It is one of the more honest single numbers on a strategy report, because unlike total return or win rate it is hard to flatter.
is a high sharpe ratio over few trades meaningful: Enough that the result cannot be explained by luck — a 2.6 Sharpe over 34 trades sits comfortably inside what randomness produces, while the same figure across several hundred trades is a genuine claim. Sample size is the precondition for every other metric mattering, and it is the number most often left out of a report. A useful companion test is to delete the best 5% of trades and rerun: a real edge degrades gracefully, a curve-fitted one collapses.
what is a good sharpe ratio for a trading strategy: There is no official threshold — above 1 is commonly described as good and above 2 as excellent, but the figure is meaningless without the trade count behind it. Sharpe measures return per unit of volatility and penalises upside volatility exactly as heavily as downside, which quietly flatters strategies that sell tail risk. Read it beside maximum drawdown and sample size rather than on its own.
why do Indian equity backtests starting in 2003 look so good: Because 2003 and 2013 both sit near the bottom of long bull runs, so almost any long-biased strategy tested from those dates looks excellent. Ask to see 2008, 2011, 2018 and 2020 broken out separately — a report showing only the aggregate is hiding the periods that decide whether the system is survivable. Live drawdowns also tend to run worse than backtested ones, because slippage widens precisely when markets move fastest.