Key points
- Golden crosses, MACD buy signals and crossing above the 200-day line did about as well as buying on any day, for every group.
- "Buy when above the 200-day line" did not work either. For individual stocks, days below the line were followed by slightly better returns.
- What held up was oversold (RSI 30 or below). For Japanese value stocks after they were picked, three-month returns beat the average by +2.7%, and were positive 68% of the time (61% for any day).
- For gold and silver no signal pointed the same way in both halves of the period. Picking the day made no clear difference.
How we tested
Buy at the close on the day of the signal and measure the return one, three and six months later (20, 60 and 120 trading days). The figures are the difference from the average return for buying on any day, same stock, same period. Repeat signals within 20 trading days count once. Splitting the period in two checks whether a result is chance.
- Gold and silver: Gold and silver futures and the gold ETF (GC=F, GLD, SI=F) (3)
- Gold miners: GDX, GDXJ and the main gold miners and royalty companies (20)
- Japanese value stocks: Companies that ranked cheap for their earnings power in their 2021–22 annual reports, with no warning signs (488)
- Japanese stocks overall: 400 listed companies picked at random (381)
- US stocks: 200 companies picked at random from today's S&P 500 (198)
Daily prices from Yahoo Finance (adjusted for dividends and splits), to 28 Sep 2026.
At a glance: three-month difference by signal and group
| Signal | Gold and silver | Gold miners | Japanese value stocks | Japanese stocks overall | US stocks |
|---|---|---|---|---|---|
| Above the 200-day line (every day) | +0.5% | −0.8% | −0.4% | −0.1% | −0.6% |
| Below the 200-day line (every day) | −0.7% | +0.9% | +0.7% | +0.1% | +1.4% |
| Crosses above the 200-day line | −2.0% | −0.7% | +0.3% | −0.2% | −0.3% |
| Golden cross (50/200-day) | −1.5% | −3.2% | +0.3% | +0.3% | −0.6% |
| Golden cross (25/75-day) | −1.8% | −2.8% | +0.1% | −0.1% | −0.7% |
| MACD buy signal (all) | +0.4% | −0.5% | −0.1% | +0.1% | −0.1% |
| MACD buy signal below zero | −0.1% | +0.9% | +0.2% | +0.2% | +0.6% |
| MACD buy signal above the 200-day line | +0.7% | −2.2% | −0.4% | 0.0% | −0.8% |
| Dip in an uptrend (RSI 40 or below, above the 200-day line) | +1.4% | −0.7% | 0.0% | −0.3% | +0.1% |
| Oversold (RSI 30 or below) | −0.7% | +3.8% | +2.7% | +3.2% | +1.7% |
| 30% or more below the 52-week high | −0.8% | +1.4% | +1.7% | +1.8% | +4.7% |
| Crosses the 50-day line after a 20% fall | −3.3% | −0.7% | 0.0% | −0.6% | +1.1% |
| Overheated (RSI 75 or above) | +1.2% | −1.9% | −1.4% | −0.4% | −1.6% |
| Overheated (1.3x the 200-day line) | +7.1% | +0.9% | −3.0% | −2.0% | +0.2% |
| Overheated (either) | +1.4% | +0.3% | −1.7% | −1.0% | −1.0% |
Difference from buying on any day. Boxed: the same direction in both halves of the period (downwards for overheated) and at least 0.5%. Shaded rows are the two signals this site uses.
Three-month difference by signal
Full figures by group
Tap to open: one- and six-month differences, how often prices rose, split periods and counts.
Gold and silverGold and silver futures and the gold ETF (GC=F, GLD, SI=F) (3, All years).
| Signal | 1 month | 3 months | 6 months | Up after 3 months (normally 56%) | Stocks pointing the same way | 2011–18: 3 months | 2019 on: 3 months | Count |
|---|---|---|---|---|---|---|---|---|
| Above the 200-day line (every day) | +0.1% | +0.5% | +1.3% | 56% | — | −0.7% | −0.2% | 6,854 |
| Below the 200-day line (every day) | −0.1% | −0.7% | −1.8% | 56% | — | +0.5% | +0.6% | 4,836 |
| Crosses above the 200-day line | −0.7% | −2.0% | −3.1% | 46% | — | −2.5% | −1.0% | 108 |
| Golden cross (50/200-day) | +0.2% | −1.5% | −1.4% | 39% | — | −3.2% | −1.2% | 33 |
| Golden cross (25/75-day) | −0.8% | −1.8% | −2.5% | 45% | — | −2.4% | −0.4% | 83 |
| MACD buy signal (all) | −0.2% | +0.4% | +0.9% | 60% | — | +0.2% | +0.4% | 316 |
| MACD buy signal below zero | −0.2% | −0.1% | −0.8% | 60% | — | +0.2% | +0.2% | 211 |
| MACD buy signal above the 200-day line | −0.2% | +0.7% | +2.2% | 59% | — | −0.2% | −0.2% | 190 |
| Dip in an uptrend (RSI 40 or below, above the 200-day line) | +1.3% | +1.4% | −1.0% | 60% | — | +3.1% | −0.8% | 87 |
| Oversold (RSI 30 or below) | +0.8% | −0.7% | −1.4% | 61% | — | −0.2% | +0.9% | 85 |
| 30% or more below the 52-week high | +1.7% | −0.8% | −1.8% | 51% | — | +2.5% | −4.1% | 37 |
| Crosses the 50-day line after a 20% fall | −1.2% | −3.3% | −5.1% | 43% | — | −2.3% | −2.0% | 53 |
| Overheated (RSI 75 or above) | +0.9% | +1.2% | +1.2% | 61% | — | −1.1% | +1.4% | 67 |
| Overheated (1.3x the 200-day line) | +1.1% | +7.1% | −0.5% | 55% | — | +12.8% | +3.3% | 22 |
| Overheated (either) | +0.3% | +1.4% | +1.2% | 58% | — | +1.6% | 0.0% | 79 |
Gold minersGDX, GDXJ and the main gold miners and royalty companies (20, All years).
| Signal | 1 month | 3 months | 6 months | Up after 3 months (normally 53%) | Stocks pointing the same way | 2011–18: 3 months | 2019 on: 3 months | Count |
|---|---|---|---|---|---|---|---|---|
| Above the 200-day line (every day) | −0.5% | −0.8% | −0.2% | 51% | 25% | −2.7% | −1.4% | 39,462 |
| Below the 200-day line (every day) | +0.6% | +0.9% | +0.2% | 55% | 75% | +1.9% | +2.8% | 34,231 |
| Crosses above the 200-day line | −0.1% | −0.7% | +0.2% | 49% | 25% | −0.5% | −0.7% | 703 |
| Golden cross (50/200-day) | −1.9% | −3.2% | −3.7% | 50% | 15% | −2.4% | −3.9% | 235 |
| Golden cross (25/75-day) | +0.1% | −2.8% | −1.1% | 46% | 5% | −2.9% | −2.2% | 565 |
| MACD buy signal (all) | −0.3% | −0.5% | −0.1% | 53% | 15% | 0.0% | −1.2% | 1,990 |
| MACD buy signal below zero | +0.9% | +0.9% | +1.1% | 56% | 70% | +1.4% | +0.9% | 1,296 |
| MACD buy signal above the 200-day line | −1.9% | −2.2% | −0.9% | 50% | 10% | −3.7% | −3.2% | 1,103 |
| Dip in an uptrend (RSI 40 or below, above the 200-day line) | +0.1% | −0.7% | −0.3% | 50% | 45% | −2.8% | −1.5% | 526 |
| Oversold (RSI 30 or below) | +3.6% | +3.8% | +3.3% | 57% | 85% | +3.7% | +6.3% | 566 |
| 30% or more below the 52-week high | +1.2% | +1.4% | +1.4% | 54% | 75% | +2.9% | +2.7% | 1,216 |
| Crosses the 50-day line after a 20% fall | 0.0% | −0.7% | 0.0% | 51% | 40% | −1.2% | +0.7% | 1,061 |
| Overheated (RSI 75 or above) | −1.1% | −1.9% | +0.3% | 47% | 70% | −0.7% | −4.6% | 359 |
| Overheated (1.3x the 200-day line) | −0.1% | +0.9% | +0.6% | 54% | 40% | +1.1% | −2.3% | 645 |
| Overheated (either) | −0.5% | +0.3% | +0.5% | 53% | 45% | +0.5% | −2.5% | 810 |
Japanese value stocksCompanies that ranked cheap for their earnings power in their 2021–22 annual reports, with no warning signs (488, After selection (Jul 2022 on)).
| Signal | 1 month | 3 months | 6 months | Up after 3 months (normally 61%) | Stocks pointing the same way | Before selection, 2011–18: 3 months | Before selection, 2019–Jun 2022: 3 months | Count |
|---|---|---|---|---|---|---|---|---|
| Above the 200-day line (every day) | −0.2% | −0.4% | −0.8% | 61% | 36% | +0.4% | −1.9% | 318,212 |
| Below the 200-day line (every day) | +0.4% | +0.7% | +1.6% | 62% | 64% | −0.7% | +2.0% | 156,430 |
| Crosses above the 200-day line | 0.0% | +0.3% | +0.7% | 61% | 50% | −0.4% | +0.2% | 4,340 |
| Golden cross (50/200-day) | +0.2% | +0.3% | +1.6% | 61% | 48% | +0.3% | +0.1% | 1,367 |
| Golden cross (25/75-day) | −0.4% | +0.1% | +0.2% | 61% | 53% | +0.2% | 0.0% | 3,393 |
| MACD buy signal (all) | −0.3% | −0.1% | 0.0% | 61% | 47% | +0.2% | +0.1% | 12,903 |
| MACD buy signal below zero | −0.2% | +0.2% | +0.8% | 61% | 55% | −0.7% | +1.4% | 7,407 |
| MACD buy signal above the 200-day line | −0.4% | −0.4% | −0.7% | 60% | 42% | +0.7% | −1.9% | 8,826 |
| Dip in an uptrend (RSI 40 or below, above the 200-day line) | +0.7% | 0.0% | +0.4% | 62% | 48% | −0.1% | −2.1% | 3,714 |
| Oversold (RSI 30 or below) | +3.1% | +2.7% | +4.8% | 68% | 72% | +0.7% | +3.2% | 2,694 |
| 30% or more below the 52-week high | +1.7% | +1.7% | +2.3% | 58% | 69% | −0.7% | +3.4% | 2,611 |
| Crosses the 50-day line after a 20% fall | −0.3% | 0.0% | +0.6% | 58% | 49% | −1.4% | +0.4% | 3,838 |
| Overheated (RSI 75 or above) | −0.6% | −1.4% | −1.7% | 59% | 68% | +1.4% | −3.1% | 3,139 |
| Overheated (1.3x the 200-day line) | −1.3% | −3.0% | −5.6% | 53% | 77% | −0.7% | −6.9% | 2,282 |
| Overheated (either) | −0.6% | −1.7% | −2.9% | 57% | 75% | +0.2% | −4.4% | 4,420 |
Japanese stocks overall400 listed companies picked at random (381, All years).
| Signal | 1 month | 3 months | 6 months | Up after 3 months (normally 54%) | Stocks pointing the same way | 2011–18: 3 months | 2019 on: 3 months | Count |
|---|---|---|---|---|---|---|---|---|
| Above the 200-day line (every day) | −0.1% | −0.1% | −0.3% | 54% | 44% | −0.1% | −0.5% | 710,588 |
| Below the 200-day line (every day) | +0.1% | +0.1% | +0.3% | 54% | 56% | +0.2% | +0.6% | 550,488 |
| Crosses above the 200-day line | −0.2% | −0.2% | +0.3% | 53% | 45% | 0.0% | −0.2% | 12,383 |
| Golden cross (50/200-day) | +0.1% | +0.3% | +0.3% | 55% | 51% | +0.9% | −0.2% | 3,892 |
| Golden cross (25/75-day) | −0.3% | −0.1% | −0.4% | 52% | 48% | +0.2% | −0.3% | 9,542 |
| MACD buy signal (all) | 0.0% | +0.1% | −0.1% | 54% | 51% | +0.2% | 0.0% | 33,495 |
| MACD buy signal below zero | 0.0% | +0.2% | +0.2% | 55% | 56% | +0.2% | +0.4% | 21,219 |
| MACD buy signal above the 200-day line | 0.0% | 0.0% | −0.3% | 54% | 46% | +0.3% | −0.6% | 19,582 |
| Dip in an uptrend (RSI 40 or below, above the 200-day line) | +0.1% | −0.3% | −0.1% | 56% | 46% | −0.4% | −0.4% | 8,700 |
| Oversold (RSI 30 or below) | +3.2% | +3.2% | +3.5% | 59% | 68% | +4.9% | +2.0% | 9,197 |
| 30% or more below the 52-week high | +1.6% | +1.8% | +2.8% | 52% | 62% | +2.9% | +1.2% | 15,649 |
| Crosses the 50-day line after a 20% fall | −0.3% | −0.6% | −0.2% | 50% | 38% | −0.7% | −0.4% | 14,421 |
| Overheated (RSI 75 or above) | −0.2% | −0.4% | −0.5% | 49% | 60% | 0.0% | −1.7% | 7,773 |
| Overheated (1.3x the 200-day line) | −0.8% | −2.0% | −3.8% | 45% | 70% | −3.7% | −1.9% | 6,475 |
| Overheated (either) | −0.4% | −1.0% | −1.7% | 48% | 64% | −1.6% | −1.6% | 11,506 |
US stocks200 companies picked at random from today's S&P 500 (198, All years).
| Signal | 1 month | 3 months | 6 months | Up after 3 months (normally 63%) | Stocks pointing the same way | 2011–18: 3 months | 2019 on: 3 months | Count |
|---|---|---|---|---|---|---|---|---|
| Above the 200-day line (every day) | −0.3% | −0.6% | −0.8% | 63% | 19% | −0.6% | −0.8% | 500,217 |
| Below the 200-day line (every day) | +0.6% | +1.4% | +1.7% | 64% | 81% | +1.6% | +1.5% | 229,276 |
| Crosses above the 200-day line | −0.1% | −0.3% | +0.2% | 62% | 48% | +0.2% | −0.5% | 6,168 |
| Golden cross (50/200-day) | −0.2% | −0.6% | +0.1% | 61% | 44% | −0.4% | −0.7% | 2,033 |
| Golden cross (25/75-day) | −0.7% | −0.7% | −0.9% | 61% | 35% | −0.3% | −1.1% | 5,329 |
| MACD buy signal (all) | 0.0% | −0.1% | −0.1% | 63% | 43% | −0.1% | −0.1% | 19,346 |
| MACD buy signal below zero | +0.4% | +0.6% | +0.7% | 64% | 73% | +0.5% | +0.9% | 10,756 |
| MACD buy signal above the 200-day line | −0.4% | −0.8% | −0.8% | 63% | 26% | −0.7% | −1.0% | 13,470 |
| Dip in an uptrend (RSI 40 or below, above the 200-day line) | +0.4% | +0.1% | −0.2% | 63% | 49% | 0.0% | +0.2% | 6,590 |
| Oversold (RSI 30 or below) | +0.5% | +1.7% | +1.6% | 65% | 77% | +1.3% | +2.5% | 4,582 |
| 30% or more below the 52-week high | +2.0% | +4.7% | +6.8% | 67% | 87% | +5.0% | +4.9% | 3,957 |
| Crosses the 50-day line after a 20% fall | +0.9% | +1.1% | +2.2% | 63% | 65% | +2.0% | +0.7% | 4,817 |
| Overheated (RSI 75 or above) | −0.9% | −1.6% | −2.1% | 60% | 76% | −1.3% | −2.2% | 4,133 |
| Overheated (1.3x the 200-day line) | −0.4% | +0.2% | +1.1% | 61% | 61% | −2.2% | +0.5% | 2,191 |
| Overheated (either) | −0.7% | −1.0% | −0.8% | 60% | 74% | −1.4% | −1.1% | 5,448 |
Stocks pointing the same way: the share of stocks whose own average went the same way as the table (downwards for overheated). It shows whether a result comes from a few stocks or many. Daily states (above/below the 200-day line) count every day.
What the test cannot tell you
- Only survivors
Yahoo has no data for delisted companies, so buying after a fall looks better than it was: the companies that kept falling are missing. The bias is largest for US stocks (today's S&P 500 members are the winners). It is smallest for the Japanese value stocks picked in 2022, from July 2022 on.
- The edge is small
The oversold edge is about 2–4% over three months on average: slightly better odds, with no lower chance of a fall of 20% or more. Costs are not included.
- Fifteen years of the past
Prices since 2011 cover low rates and the rate rises, but the future may differ. Widely known signals can stop working.
- Not the site's own list
TOB candidates were tested on Japanese stocks overall and value stocks on Japanese value stocks picked in 2022. Today's lists may behave differently.
Updated 29 Sep 2026 02:37 JST · closing prices on 28 Sep 2026