Methodology and data
How the list is made: where the numbers come from, when they are calculated and where their limits lie.
A ranking is only as good as its data. This page sets out what the RSL list is based on — including where the answer is not flattering. How the RSL itself is calculated is covered in RSL: Method & Origins, how to read the table in How to read the RSL list.
Where the data comes from
Two kinds of data: the prices from which every number is calculated, and the lists of which stocks belong to an index.
Prices
The closing prices come from Yahoo Finance (yfinance). For each stock the list loads about 2 years of daily closing prices and calculates every value itself; no ready-made figure is taken over.
To be honest: this source is free and unofficial. It makes no promise about availability or completeness, and its terms do not cover commercial use. This site therefore serves no advertising. A licensed source can be connected without rebuilding anything, but has not been chosen yet.
Index constituents
Which stocks belong to an index comes from Wikipedia and is published there under the CC BY-SA 4.0 license. The list takes the symbols, names and sectors from it and rewrites the symbols into the notation of the price source — an excerpt, edited. The constituents are loaded afresh with every run; when a stock changes index, the list follows as soon as Wikipedia has recorded the change.
| Index | Constituents |
|---|---|
| S&P 500 | Wikipedia, afresh with every run |
| NASDAQ 100 | Wikipedia, afresh with every run |
| DAX 40 | Wikipedia, afresh with every run |
| Dow Jones Industrial | stored in the code, 30 stocks, as of 2026-09-18 — not yet cross-checked |
| DJ Transport | Wikipedia, afresh with every run |
| DJ Utility | Wikipedia, afresh with every run |
For the Dow Jones Industrial, Wikipedia no longer has a machine-readable table of constituents. Its 30 stocks are therefore stored in the code. That is acceptable because this index rarely changes — but such a list goes stale silently if nobody maintains it. The list as of 2026-09-18 has not yet been checked against an official listing.
When the list is calculated
Once a day, at 22:30 German time — shortly after the close in New York.
A run loads the constituents of all 6 indices and the prices of their stocks, and recalculates the whole list from them. The numbers on the site therefore apply to the closing prices of the trading day in question, not to the moment you read them. When the last run took place is shown above every list.
Whether an exchange traded on a given day is read from the data itself, not from a holiday calendar: if the latest price of an index is from the same day, its exchange traded. The US indices thus follow the trading days of the US exchanges, the DAX those of the Frankfurt exchange. Report emails go out only after a trading day.
Known limits
What the list cannot do belongs here just as much as what it can.
Only today's members of an index
The list works with the stocks that belong to an index today. Anything that has left the index — often after a weak period — no longer appears. For a ranking of today that is correct. But anyone who wanted to read from it how well a selection would have worked in the past would see only the survivors. Research calls this survivorship bias.
Missing prices
If the source delivers no prices for a stock, or it has fewer than 27 weekly closing prices, it is missing from that day's list — better no value at all than an average over too few prices. If a whole index turns out much smaller than usual or its latest price is old, the monitoring flags the run.
Differences from other RSL lists
This list calculates Levy's RSL with 27 weekly closing prices in the denominator: the current one and the 26 before it. At least one other widely used list averages only 26. On 2026-09-20 we compared 10 values with the export of another RSL list. Calculated with its 26 prices, all matched to 4 decimal places. If a value here differs from another list, that may be due to the definition, a different reference date or a different price source.
Source: Robert A. Levy 1967, Section II (“Price Ratios”)
What is expressly not done
No backtest of our own, no tuning of parameters.
This list does not calculate any return that could have been achieved with it in the past, and it shows none. Such a backtest on a few years and a few hundred stocks almost always finds a setting that looks good — and has found nothing by doing so. That is exactly what Jensen and Benington held against Levy in 1970: in his dissertation he had tested some 68 variations of trading rules on the same data.
The windows and parameters of the measures are therefore exactly as the literature gives them. None has been tuned, including those of the four research columns. The reason for choosing them is the evidence behind them, not how they perform on our data. More on this in The setback and Why the RSL stays what it is.
Source: Michael C. Jensen 1970, passage on Levy's dissertation (1966)
Not investment advice and not a recommendation. Past performance is not a reliable indicator of future results, and investing in securities involves the risk of losing your entire investment.
References
Every figure on this page is shown with its exact location in the source — section, table or page. The links lead to the paper via its DOI; the full texts are mostly behind publishers' paywalls, but the bibliographic details are enough for any library.
- Robert A. Levy (1967): Relative Strength as a Criterion for Investment Selection. The Journal of Finance 22(4), pp. 595–610. doi:10.1111/j.1540-6261.1967.tb00295.x
- Michael C. Jensen, George A. Benington (1970): Random Walks and Technical Theories: Some Additional Evidence. The Journal of Finance 25(2), pp. 469–482. doi:10.1111/j.1540-6261.1970.tb00671.x