Rebound: how the watchlist is built and what it does not promise
A rebound is a recovery after a sharp fall in price. The Rebound page shows which S&P 500 stocks are currently lagging well behind the market, and what that decline looks like. It is a watch list, not a buy signal.
Important: The list does not say which stock will recover. The backtest showed no reliable edge for any of the features (section 5). That is why the page deliberately has no grade and no chance of success.
1. How the selection works
- Stock list: the roughly 500 members of the S&P 500. The member list is reloaded at most every seven days, the prices daily.
- Lag behind the market: The stock's return over the last 10 trading days minus the return of the MSCI World (ETF URTH) over the same period. Only stocks that are at least 7 percentage points behind make the list.
- Turnover filter: The 30 % of stocks with the lowest average trading turnover (50 days) are dropped. A stock that is hardly traded is difficult to get into and out of.
- Recency and history: A stock is not shown without a price from the last 5 days or with fewer than 260 trading days of history.
- Sorting: by the lag, largest first. At most the 60 strongest stocks are shown as cards; the line below the heading gives the total number.
The list is rebuilt once a day with the daily routine. "As of" is the last trading day whose prices were included.
2. Reading a card
| Row | Meaning |
|---|---|
| Drop 10 d | Lag behind the market over 10 trading days. The bar is full at 25 %. |
| Oversold | RSI with period 2: a very short-term momentum value from 0 to 100. The lower, the more oversold; the bar is full at RSI 0. More on the RSI in the article RSI. |
| Trend | Distance of the price from the 200-day line. In the middle of the bar the price is on the line, to the left below it, to the right above it. |
| vs. sector | Lag behind the market minus the average lag of the sector (GICS sector) among the S&P 500 stocks. If almost the whole sector has fallen, the value is small. Shown when the sector has at least 4 stocks. |
| calm / News? | An approximation, not a real news check: News? appears when, in the last 10 days, there was a price gap or a daily loss of at least 3 ATR, or a falling day with at least three times the usual volume. If a stock falls without such a trigger, it shows calm. |
| Stop 2 ATR | Distance of a stop of two average daily fluctuations (ATR over 14 days), in percent of the price. More on this in the article ATR stop. |
The bars show the strength of the feature, not the stock's prospects. A full bar means "fallen sharply", not "will rise".
3. The band at the top and the filters
- VIX: the volatility figure of the US market. From 25 upwards the environment counts as stress, and the band gets a red border. After stress, the recovery in the backtest was somewhat stronger, but not robust.
- Stocks above the 200-day line: Share of the checked S&P 500 stocks with a price above the line. It shows whether the overall market is more in an uptrend or a downtrend.
- Filters: all, without suspected news, above the 200-day line, below the 200-day line. They only hide cards; the selection stays the same.
4. Why the page has no score
Originally, an overall score made up of lag, oversold depth, quality and trend was planned. Before it went onto the page, it was tested on price data. The result is in the next section. A score without proven effect would feign an accuracy that does not exist.
5. What the backtest showed
All 499 stocks with sufficient history were tested at every weekend from November 2019 to November 2026 (365 weeks, around 179,000 data records), with results after 5, 10, 21 and 63 trading days compared with the MSCI World.
- No feature has a robust effect: Lag, RSI(2), decline in ATR, position relative to the 200-day line and distance to the 52-week high never have a t-value above 2. Only a t-value of 3 or more counts as a finding, with the same sign in both halves of the period.
- Nor does the overall score: The top fifth performed hardly any differently from the bottom fifth. After assumed costs of 0.4 % per round trip, about zero or less remains.
- News filter: The difference between "calm" and "News?" was not robust.
- Market environment: With a high VIX, the top fifth was about 0.7 percentage points ahead of the average after 21 days (t ≈ 1.1). A hint, nothing more: the stress weeks come from a few episodes.
- Sector adjustment: This made the relationship weaker. So part of the lag is just a sector move.
- Only the most extreme cases: The top 5 % were about 0.5 to 0.6 percentage points ahead of the average after 21 days (t about 1.0 to 1.2), likewise not robust.
- Falling knife: Around a quarter of all stocks were more than 5 % behind the market after 21 days. For stocks with a strong lag, it was no fewer.
For comparison, the page shows at the bottom how similar cases turned out in the past, including the t-value. More on the backtest method is in the chapter Backtest.
6. Limits
- Only US stocks from the S&P 500, no European stocks.
- Survivorship Bias: Today's members were tested. Stocks that have left the index are missing, so the results tend to be too good.
- No quality check: Balance sheet data is only available as it stands today, so a quality filter could not be backtested. The list does not distinguish between a healthy company with a price decline and a company in difficulties.
- News only approximated: A price gap or a volume surge is no proof of a news item, and a calm decline can still be justified.
- Many tests: With so many comparisons, individual conspicuous values arise by chance alone. Hence the strict threshold of t = 3.
- Look back: Everything describes the past. Costs, thresholds and periods are assumptions.
How to use the list sensibly
As a starting point for your own research: why has the stock fallen, is there a news item, does the decline fit the sector, and how far away is the stop? The page does not do this check for you.
Education, not investment advice. No stock on the list is a recommendation, and a sharp decline can continue.
Education, not investment advice. · Next: Backtest: what two backtests show and what they don't · ATR stop: how far from the price? · RSI: overbought, oversold, or both wrong? · Glossary