Claim check: Buy the dip
“Buy the dip: after a down day or pullback, prices bounce.”
What the data says
Partly Partly: after “5-day fall of 3% or more, above the 200-day average”, the average 21-day return was +1.0% against +0.7% for all days (medians over 9 markets); it was higher than usual in 8 of 9 markets and statistically clear, with at least 20 independent events and the same sign in both halves, in 1. The direction mostly points the claimed way, but not strongly enough to be sure.
It holds in some markets or periods, leans the right way without proof, or helps on risk but not on return.
Who says it
Retail rallying cry on Reddit's WallStreetBets and everywhere in the 2010-2021 bull market.
Origins are as commonly cited and were not independently verified.
How we tested it
Event: the close is down 3% or more over five trading days while it is above its 200-day average (a dip inside an uptrend). We record the 21-trading-day return from that close (events at least 21 days apart) and compare it with the average 21-day return of the same market. The same fall below the 200-day average is shown for contrast and does not count towards the verdict.
Results
5-day fall of 3% or more, above the 200-day average 21-day (claim: higher than average)
| Market | Events | Mean after | All days | Difference | Median after | Up share | t-stat | Reading |
|---|---|---|---|---|---|---|---|---|
| S&P 500 | 38 | +2.1% | +1.2% | +0.9% | +3.1% | 71% | 0.9 (likely noise) | Right way, likely noise |
| Gold | 41 | +1.0% | +0.7% | +0.3% | +1.0% | 56% | 0.4 (likely noise) | Right way, likely noise |
| Silver | 56 | +0.9% | +0.9% | 0.0% | +0.2% | 50% | 0.0 (likely noise) | Right way, likely noise |
| Oil | 60 | +0.3% | +0.3% | 0.0% | -0.7% | 47% | 0.0 (likely noise) | Right way, likely noise |
| Natural gas | 40 | +0.3% | -1.5% | +1.9% | -2.6% | 42% | 0.6 (likely noise) | Right way, likely noise |
| Commodities | 39 | +2.1% | +0.4% | +1.7% | +2.5% | 74% | 2.2 | Clear and stable |
| Agriculture | 24 | +0.1% | +0.2% | 0.0% | -0.1% | 46% | 0.0 (likely noise) | No effect |
| EUR/USD | 4 | -0.9% | -0.1% | -0.8% | -0.5% | 50% | -0.4 (likely noise) | Too little data |
| Bitcoin | 82 | +5.1% | +4.0% | +1.1% | +1.8% | 59% | 0.5 (likely noise) | Right way, likely noise |
| Ether | 56 | +4.4% | +3.6% | +0.8% | +0.7% | 54% | 0.2 (likely noise) | Right way, likely noise |
Difference from the average, by market (21-day)
By horizon: mean difference from all days
| Market | 5-day | 21-day | 63-day |
|---|---|---|---|
| S&P 500 | +0.3% (48) | +0.9% (38) | +0.8% (27) |
| Gold | +0.5% (56) | +0.3% (41) | +1.6% (27) |
| Silver | -0.2% (101) | 0.0% (56) | +1.8% (30) |
| Oil | -0.2% (100) | 0.0% (60) | 0.0% (32) |
| Natural gas | -0.5% (77) | +1.9% (40) | +4.1% (18) |
| Commodities | 0.0% (53) | +1.7% (39) | +3.0% (22) |
| Agriculture | 0.0% (34) | 0.0% (24) | +0.1% (16) |
| EUR/USD | -0.4% (5) | -0.8% (4) | -1.9% (3) |
| Bitcoin | -0.3% (175) | +1.1% (82) | +4.2% (39) |
| Ether | +0.7% (128) | +0.8% (56) | +7.9% (24) |
Average -day return after the signal vs all days (S&P 500)
Same fall, below the 200-day average · For comparison 21-day (claim: higher than average)
| Market | Events | Mean after | All days | Difference | Median after | Up share | t-stat | Reading |
|---|---|---|---|---|---|---|---|---|
| S&P 500 | 30 | +1.8% | +1.2% | +0.6% | +1.6% | 53% | 0.4 (likely noise) | Right way, likely noise |
| Gold | 45 | +1.1% | +0.7% | +0.4% | +1.0% | 60% | 0.5 (likely noise) | Right way, likely noise |
| Silver | 78 | +0.4% | +0.9% | -0.5% | -0.3% | 47% | -0.6 (likely noise) | No effect |
| Oil | 89 | -0.1% | +0.3% | -0.4% | +1.8% | 61% | -0.3 (likely noise) | No effect |
| Natural gas | 133 | -1.0% | -1.5% | +0.5% | -2.2% | 45% | 0.4 (likely noise) | Right way, likely noise |
| Commodities | 53 | -0.6% | +0.4% | -0.9% | -0.2% | 47% | -1.2 (likely noise) | No effect |
| Agriculture | 34 | -0.4% | +0.2% | -0.5% | -0.6% | 44% | -0.9 (likely noise) | No effect |
| EUR/USD | 17 | +1.3% | -0.1% | +1.4% | +1.0% | 76% | 2.7 (likely noise) | Right way, likely noise |
| Bitcoin | 68 | +1.7% | +4.0% | -2.4% | +1.1% | 57% | -1.3 (likely noise) | No effect |
| Ether | 62 | +0.5% | +3.6% | -3.1% | -3.8% | 40% | -1.1 (likely noise) | No effect |
9 of 10 markets had enough data.
Stability across the two halves of history
A saying that only works in one half of the history is not much of a rule. Each row shows the effect in the full history, the first half and the second half; green means the claimed effect was present.
| Market | Full history | First half | Second half |
|---|---|---|---|
| S&P 500 | +0.9% (38) | +1.1% (16) | +0.7% (22) |
| Gold | +0.3% (41) | -0.7% (13) | +0.5% (28) |
| Silver | 0.0% (56) | -0.6% (20) | 0.0% (36) |
| Oil | 0.0% (60) | +0.3% (25) | -0.5% (35) |
| Natural gas | +1.9% (40) | -1.1% (17) | +3.9% (23) |
| Commodities | +1.7% (39) | +3.1% (14) | +0.7% (25) |
| Agriculture | 0.0% (24) | +0.7% (6) | -0.6% (18) |
| EUR/USD | -0.8% (4) | -0.8% (4) | — |
| Bitcoin | +1.1% (82) | +0.8% (43) | +1.3% (39) |
| Ether | +0.8% (56) | +1.0% (29) | +0.4% (27) |
The honest caveat
Dips in a long bull market look good partly because the whole period was a bull market.
Descriptive statistics from historical data, before costs. Not a forecast and not investment advice.
Questions
Buy the dip: after a down day or pullback, prices bounce.
Partly: after “5-day fall of 3% or more, above the 200-day average”, the average 21-day return was +1.0% against +0.7% for all days (medians over 9 markets); it was higher than usual in 8 of 9 markets and statistically clear, with at least 20 independent events and the same sign in both halves, in 1. The direction mostly points the claimed way, but not strongly enough to be sure.
How was this claim tested?
Event: the close is down 3% or more over five trading days while it is above its 200-day average (a dip inside an uptrend). We record the 21-trading-day return from that close (events at least 21 days apart) and compare it with the average 21-day return of the same market. The same fall below the 200-day average is shown for contrast and does not count towards the verdict.
Markets used
S&P 500 · Gold · Silver · Oil · Natural gas · Commodities · Agriculture · EUR/USD · Bitcoin · Ether