Education Centre / Reading your result
Reading your result: what each number means
13 min read · Walk-through
In everyday words
A school report has a mark for each subject and a line from the teacher. A result is the same: a few lines, each saying one thing.
An example in pounds
In the example on this page, £1,000 became £1,564 with the rule and £1,764 by simply holding. It is one past test. It is not a prediction.
The short version
Every line of a result, from the top: what it is, why it is there, and how to read it.
This page walks down a result, one line at a time. First comes the short card, which a new account opens on. Then come the main numbers of the full report.
The example is a real result from a test that was run on Quantis. It is an example of a past test, not a prediction. Your own result has the same lines, with your own figures in them.
A result from a newer test has one more line, just above the last one. It says when the trades were made. A second line shows there only when an order was large beside what was traded that day.
The person asked for a golden cross on ISF.L. ISF.L is a fund that follows the FTSE 100, a list of 100 of the largest companies on the London market. The second line of the card says what a golden cross is.
A test describes the past. It is not a forecast and it is not advice.
The short card, top to bottom
- You asked:
- Holds ISF.L while its 50-day average price is above its 200-day average, and stays in cash otherwise.
- Lost to simply holding ISF.L.
- Worst fall about the same: 13% against 13%.
- This rule
- £1,564
- from £1,000
- Simply holding
- £1,764
- from £1,000
- £1,000 put in at the close of 6 Oct 2021, counted at the close of 5 Oct 2026.
- Only 3 round-trip trades in 5 years, so one decision can swing this result.
- A backtest of 7 Oct 2021 to 5 Oct 2026. It describes the past. Not a forecast, not advice.
- What each line means
- Full report
You asked
What it is. Your own words, exactly as you typed them.
Why it is there. It shows at a glance that this is the test you meant.
How to read it. There is no good or bad here. If the words are not what you meant, the card answers a different question.
What the rule does
What it is. One sentence saying what was tested, in plain words.
Why it is there. A name like golden cross can be read in more than one way. This is the reading the test used.
How to read it. It matches what you had in mind, or it does not. If not, the result is about a different rule from yours.
Beat or lost
What it is. The headline: did the rule end with more or less than simply holding? Simply holding means buying on the first day and doing nothing after.
Why it is there. A result means little until it is set beside something. Doing nothing is the plainest thing to set it beside.
How to read it. Beat means the rule ended with more over these dates. Lost to means all the buying and selling ended with less than doing nothing. About the same means a tie. Here the rule lost. None of the three says what happens after the last day of the test.
Worst fall
What it is. The biggest drop from a high point to the low that followed. The first figure is the rule's, and the second is for simply holding.
Why it is there. It is the part of a result people feel. A test can end well and still hold a stretch that was hard to sit through.
How to read it. Smaller is easier to sit through. A fall of 13% takes £100 at a high down to £87. A fall of 50% takes it down to £50. Here the two fell about as far as each other.
The two amounts
What it is. What the starting amount became, under the rule and under simply holding.
Why it is there. Pounds are easier to picture than percentages.
How to read it. The larger amount ended ahead. Here simply holding ended £200 ahead of the rule. An amount below the starting amount means money was lost over those dates.
The dates
What it is. The day the pretend money went in and the day it was counted. At the close means at the last price of that day.
Why it is there. The same rule over other dates gives other amounts.
How to read it. A longer stretch has been through more, good years and bad. A short one can be a single lucky or unlucky spell. This one is five years.
The caution
What it is. A warning that the test had little to go on. A round-trip trade is one buy and the sell that follows it.
Why it is there. With three trades, one lucky trade can lift the whole result and one unlucky trade can sink it.
How to read it. It appears only when a test had little to go on, such as fewer than ten finished trades. A card with no caution had more to go on. That is not a sign that the rule works.
The last line
What it is. The dates that were tested, and what a test is. A backtest is a replay of a rule on past prices.
Why it is there. Every result is about the past, and this line keeps that in sight.
How to read it. It reads the same on every result. The first tested day is the day after the money went in, because the money goes in at the close of the day before.
The full report: the questions
The full report opens on a summary. When a rule made less than its benchmark, one sentence at the top says so. Then come its questions, each answered from the test's own figures.
It says benchmark where the short card says simply holding. A benchmark is the thing a rule is compared with. Its phrase for buying once and doing nothing after is buy-and-hold.
Did it beat its benchmark?
On the example
Did it beat its benchmark (FTSE 100 (ISF.L, iShares Core FTSE 100), buy-and-hold)? No (9.4% vs 12.0% a year)
What it is. The same comparison as Beat or Lost to, as growth a year. The first figure is the rule's.
Why it is there. It is what most people look for first.
How to read it. Yes means the rule grew faster than its benchmark over these dates. No means slower. About the same means a hair apart. Here the rule grew 9.4% a year and simply holding grew 12.0%.
Was it less risky?
On the example
Was it less risky? About the same (worst fall 12.7% vs 12.7%)
What it is. The two worst falls, side by side. The first is the rule's.
Why it is there. Some rules make less and also fall less. This line shows whether that happened.
How to read it. Yes means the rule's worst fall was the smaller one. No means the bigger. Here they are the same to one decimal place. The short card rounds both to 13%.
Big falls when its history is reshuffled?
On the example
Big falls when its history is reshuffled? 158 of 1,000
What it is. The test's own daily ups and downs are reshuffled 1,000 times. This counts the reshuffles that fell more than 20% at some point.
Why it is there. The order the good and bad days came in is partly luck. This shows how much the worst fall leaned on that order.
How to read it. A high count means many orderings of the same days held a big fall. A low count is not a sign of safety. Here it is 158 of 1,000, about one in six. It counts reshuffles of the past, and it is not a forecast.
Luck check: did it beat chance?
On the example
Luck check: did it beat chance? Not run
What it is. Every new test also runs the rule many more times in two ways. One buys and sells on random days, for the same time in the market. The other uses the same prices in a shuffled order. The line says whether the real rule did better than nearly all of those runs.
Why it is there. A rule that does no better than random trading days could have come out ahead by luck. When the check does not pass, the card does not say the rule beat simply holding.
How to read it. Yes means the rule beat the mark against both kinds of run. No means this could be luck. It is a lighter version of two of Prove it's attacks, and passing it still describes the past. This example was tested before the check existed, so it says Not run.
Could it be luck?
On the example
Could it be luck? Yes, it could be
What it is. After the answer comes a score from 0 to 100 for how firmly the result stands above what luck alone would be expected to show. The full report calls it the deflated Sharpe ratio.
Why it is there. Try enough versions of a rule and one looks good by luck. This line allows for the versions that were tried: 8 here.
How to read it. Under 95, the answer is Yes, it could be: luck is not ruled out. At 95 or more, the line says the result clears that bar. Here the score is 90%, so the answer is yes. A high score is still not proof that the rule works, and it is not a forecast.
Survives Prove it?
On the example
Survives Prove it? Not tested yet
What it is. Prove it is a second, harder test that can be run on a result. A set of attacks tries to break it, such as shuffling the prices and tripling the costs.
Why it is there. A result that falls apart under a small change was fragile from the start.
How to read it. Until it is run, the line says Not tested yet. After that it says how many of the attacks the result survived. More is sturdier, and surviving all of them still describes the past. It is optional, and it costs tokens.
The full report: five key numbers
Under the questions are the amounts in pounds and five key numbers. The simpler view shows three of the five, with a link to the other two.
The summary holds two more things this page does not walk through: a line on what trading cost, and a row of small labels. Each label opens a note about the test. A newer result adds a third: one sentence on when its trades were made.
The amounts in pounds
On the example
£1,000 put in at the close of 6 Oct 2021 would have been £1,564 at the close of 5 Oct 2026; the benchmark: £1,764.
Your £1,000 grew to £1,250, then fell to £1,092 at its worst (down 13%) and took 3 months to get back to £1,250.
What it is. The two amounts from the short card, in one sentence. When a result has it, a second sentence tells the worst fall in pounds.
Why it is there. The second sentence adds what a percentage leaves out: how long the fall took to make back.
How to read it. A fall made back in 3 months is a different thing to live through from one that takes 4 years. Here it took 3 months.
Growth a year
On the example
Growth a year (CAGR) 9.4%
Benchmark: 12.0%
What it is. The steady yearly growth that would have turned the starting amount into the final amount. CAGR is its short name.
Why it is there. Tests cover different lengths of time. A yearly figure lets a 3-year test sit beside a 10-year one.
How to read it. Higher is faster growth, and below zero means the money shrank. It means most beside the benchmark's figure. Here the rule's £1,000 became £1,564 in five years, which is 9.4% a year. The benchmark's became £1,764, which is 12.0% a year.
Worst fall, again
On the example
Worst fall (max drawdown) 12.7%
Benchmark: 12.7%
What it is. The same worst fall as on the short card, to one decimal place. Max drawdown is the full report's name for it.
Why it is there. It sits beside the growth figure so the two are read together: what was gained, and what had to be sat through.
How to read it. Smaller is easier to sit through. A fall of 12.7% takes £1,000 at a high down to £873.
Tested period
On the example
Tested period 5.0 years
1,259 daily bars
What it is. How long a stretch of history the test covered. A daily bar is one day of prices, so 1,259 daily bars is 1,259 trading days.
Why it is there. Every other number on the result is about this stretch and no other.
How to read it. Longer has been through more. A few calm years can pass without a deep slump, and a test over those years has not met one.
Hit rate
On the example
Hit rate (days in the market, not trades) 56%
Too few closed trades (3) for a trade win rate
What it is. Of the days the rule was holding something, the share that ended higher than they began.
Why it is there. It answers how often. It does not answer how much.
How to read it. Above 50% means more up days than down days. That is not the same as making money: a few big down days can outweigh many small up ones. Here it is 56%. The line under it is about a second figure, the win rate, which counts finished trades and not days. A closed trade is a finished one, and three are too few to work it out.
Further down the full report
The full report has more numbers further down. Each has a page of its own here.
- The Sortino ratio: what was gained for each bad bump, leaving the good surprises out.
- Turnover: how much of what was held got swapped each year.
- Three figures that ask about luck: the checks behind the question Could it be luck.
What a result does not tell you
- What happens next. A result describes the dates it covered and nothing after them.
- What to buy. A result is not advice.
- How the rule does elsewhere. This example is one rule on one fund over five years. It says nothing about other shares or other dates.
Read next
- Read your result → the same walk on your own latest result, one question at a time.
- Start here → shares, funds and tests on past prices, from zero.
- Overfitting → why a test that looks good can still be wrong.