Kuantile
Was It a Good Week? Judge Your Return by Its Historical Percentile
"My portfolio fell 2% this week" says little on its own. For a calm portfolio, 2% is a jolt; for a volatile one, a routine week. The way to give the number meaning is to compare it with your portfolio's own history — in statistical terms, to find its percentile.
What is a percentile?
Line up all of your portfolio's historical weekly returns from worst to best. Where does this week fall? At the 80th percentile you have just had a week better than 80% of your history; at the 5th you are living one of your worst. The same −2% can land at the 35th percentile in one portfolio (notable but ordinary) and the 2nd in another (alarm). Context is worth more than the number.
How Kuantile reports use it
In the weekly, monthly and yearly report emails you will find a line like: "This period's return sits in the 72nd percentile of your portfolio's historical weekly returns (above the historical average)." The computation: every historical 7-day window of your current weights is evaluated (on a calendar-day basis, so assets trading 7 days a week — like crypto — share the same window as weekday-only markets), and the current period is located within that distribution.
What percentiles capture — and what they don't
- They capture: "is this drop normal for my portfolio?" Consecutive sub-10th-percentile periods hint that your portfolio's character has changed — perhaps a recent addition broke the regime.
- They don't capture: your absolute goals. Being at the 90th percentile of your own history does not mean you beat inflation; a low-volatility portfolio's "great week" can still be modest.
The behavioural payoff: calibrated panic
Losses feel roughly twice as intense as equivalent gains, which is why an ordinary bad week can trigger panic selling. A line that says "38th percentile — near your historical average" answers emotion with a number. The reverse also holds: a run of 95th-percentile periods is usually evidence not of genius but of quietly growing risk. Good risk management means acquiring tools that blunt panic on bad days and hubris on good ones.
What exactly is a percentile?
When you sort a distribution from smallest to largest, a value's "percentile" is the proportion of observations below it. Your return being in the 10th percentile means 90% of similar past periods did better; being in the 90th means very few periods beat you. This lets you see the entire distribution that a single "average return" number hides — where your good and bad scenarios actually stand.
Why the average misleads
The average is hypersensitive to extreme values and often misrepresents the "typical" experience. If your portfolio gained 200% in one year and lost 10% in each of the other four, the average looks positive — but in reality you lost money in most years. The median (50th percentile) avoids this trap: it shows the exact middle of observations. Kuantile presents your return distribution in percentile bands so you can read across "worst 5%, median, best 5%" without fixating on a single number.
Reading percentile bands
The most powerful use of percentiles is thinking of the future as a range. Instead of "my expected return is 20%," saying "with 90% probability my return is between -15% and +55%" is far more honest and prepares you for uncertainty. These bands show the same statistical tail as VaR from a different angle: VaR gives the boundary of the lower tail, while percentile bands make the full spread of both bad and good scenarios visible. Reading your portfolio with both VaR and percentiles lets you grasp risk as both a number and a feeling.
Open your reports and find your percentile →