How we calculate hit rate

When we show a hit rate for an analyst house, we measure how the house’s published price targets have performed after the fact. Here we explain exactly how the numbers are calculated.

Hit rate (12 months)

A price target counts as hit if the share price reaches the target within 12 months of publication. For targets above the price at publication, the price must rise to the target. For targets below the price, a hit means the price falls to the target. Both directions count, because both are predictions about where the price is heading.

Direction accuracy and median error

Hit rate does not tell the whole story. That is why we also show how often the stock moved in the direction the target implied, and the median error: how far from the actual 12-month outcome the target typically was. A house that often sets high targets can achieve a high hit rate in a rising market without being more skilled than others. Direction and median error make the picture fairer.

Sample rules

Only targets that are at least 12 months old are counted. Younger targets have not had time to show their outcome.
If a house repeats the same unchanged target for the same stock within 90 days, it is counted only once.
We only show percentages for houses with at least 30 evaluable targets. With fewer observations the numbers are misleading.
Targets on stocks where we lack price history, or that were delisted shortly after publication, are not counted.

Sector strength

On each house’s page we also break the hit rate down by sector. The same counting rules apply, but a sector is only shown when the house has at least 30 evaluable targets in it, and stocks without a sector classification are excluded. Each sector is compared with the house’s own average, so you can see where the house is stronger or weaker than its own level.

Consensus and agreement

On the stock pages, consensus is the median of each house’s latest price target, counting only targets set within the last 90 days. The median keeps a single extreme target from dragging the value. Targets more than three times above or below the current price - for example targets left over from before a split - are excluded.

Agreement is measured with the coefficient of variation (standard deviation divided by the mean) over the same fresh targets: below 8% we show “Aligned”, 8-16% “Some disagreement” and above 16% “Divided”. The range (lowest-highest target) and the number of houses are always shown next to the figure. When fewer than two houses have fresh targets, we show each house’s latest target regardless of age, clearly marked as older data.

Data and limitations

The statistics are based on publicly published price targets since 2020 and daily share prices. The method measures what actually happened after publication, but does not account for dividends or for the fact that different houses cover different stocks and time periods. Prices are unadjusted daily closes, so stock splits are not adjusted for: a target set before a split may be counted as hit or missed on the wrong basis. The figures are updated daily and represent history, not a guarantee of future accuracy.

Optimism and deviation from peer houses

Beyond hit rate, we measure whether a house systematically sets higher or lower price targets than its peers. For each target we compare the house’s target with the average of the other houses’ fresh targets for the same stock (targets published within the previous 30 days, same day included). The deviation is expressed as a percentage: a target of 110 against a peer average of 100 gives +10%. Single extreme deviations beyond ±40% are capped at exactly ±40% so outliers do not dominate. The house’s figure is the median of the deviations over the last 12 months.

The median makes the measure insensitive to how often a house repeats the same target.
We only show the deviation for houses with at least 30 comparable targets, and the number of targets (n) is always shown next to the figure.
A deviation above +2% is labelled “Optimistic”, below −2% “Conservative”, and in between “In line”. It is not a grade: houses cover different stocks, and a deliberate bullish view can produce higher targets without being wrong.
Name variants of the same house are merged before comparison (e.g. “SEB” and “SEB Equities” count as the same house), so a house is never compared with itself.