Help · Methodology

Classic and Peer Screener scores

From RangefinderInvest's built-in help · applies to version 0.59.3

The Screener keeps two scoring methods side by side. Both summarize stored history to help build a shortlist. Neither predicts next year's winner or recommends what to buy.

Classic score: broad comparison

Classic score is the lens for a broad universe or a view containing more than one category. It blends trailing return, subtracts an explicit expense-ratio and risk-free-rate drag, scales the result by total volatility, then applies drawdown, calendar-consistency and return-concentration penalties. Confidence shrinks a fund with a short record, measured against the history the app has stored for it up to the end date, which the Days of history column shows: at an earlier end date, a price dated after it does not count. The exact weights and floors are in the Screener's own Scoring rules drawer (opened beside the ranked results) and take effect immediately.

A fund with no expense ratio on file is charged the median of its category's known expense ratios, not 0%, and its ER cell shows that figure with a leading "~" and a tooltip saying it is an estimate. Charging an unknown fee as zero let a fund with no ER on file outrank the same fund at 0.03%. (A category with no known expense ratio has nothing to estimate from; its funds with no ER are charged 0% as before.)

Funds of one category are measured to the same end date. When a mutual fund's NAV for today has not posted but its ETF twin's close has, the category is measured to the last day all of its currently trading members have, so identical holdings read the same year-to-date and one-year return. A fund silent for more than ten days is not waited for.

Classic score is the older whole-universe measure. It is useful for sorting a broad result set, but different asset classes still carry different return and risk structures; read the category and risk columns beside it.

Peer score: one category only

Peer score asks how an equity fund compared with others in the app's same category. It blends de-overlapped return segments, divides by downside deviation, and standardizes that risk-adjusted return using the category's median and median absolute deviation. Share classes of one fund, and index funds that track the same index, count once in that median and spread. These are funds whose yearly returns have differed by the same fee-sized amount for at least five years. Each fund is still scored on its own return, so a cheaper class edges out a dearer one by its fee difference and no more. Two gates then shrink that centered score toward neutral: confidence, from how much of the record exists (both the length of the stored history and how many of the return segments it reaches), and concentration, when one calendar year carried the fund's whole gain by more than the number of its up years would ordinarily produce. A fund with an ordinary spread of good years takes no concentration haircut, however long or short its record. Concentration lowers the trust placed in a record, and a gate shrinks a score toward 0, the category's centre (its median, or its mean when robust standardization is off): above the centre that lowers a fund's score, and below it the score moves toward 0, which is higher than an otherwise identical steady record would get. The Shortlist flag marks the configured top N within each category.

The optional manager-skill weight in Scoring rules compares a fund's calendar years with its category's benchmark index. It needs at least five years the two have in common. A fund with fewer has no skill term and is scored exactly as it would be with that weight at zero.

There is no separate fee term: adjusted-close returns are already net of the fund's expenses, so subtracting the expense ratio again would double-count it. There are also no separate total-volatility, max-drawdown or calendar- consistency penalties. Non-equity and uncategorized rows show a dash because this lens was not designed for them.

Peer score is category-relative. A 0.4 in Emerging Markets is not comparable with a 0.4 in U.S. Large. The Screener only uses it as the ranking lens when the current scope contains one category. Use either score to shortlist, then review the fund, its fit and the underlying history.