PLTR daily-return tails over the last ~3 years: upside melt-ups vs downside crashes
PLTR’s largest single-day shocks over the past roughly three years have tended to be melt-ups, not crashes. Measuring close-to-close returns across 751 trading days, the return series shows a positive Fisher-Pearson skew of about +0.42; the biggest up day was +24.51% while the largest down day was −15.08%. That tilt shows up in the extremes: a majority of the largest absolute moves and a slight edge in |return| ≥ 5% days are on the upside.
The dataset and methods are straightforward — resampled minute bars to daily closes, 36-month trailing returns, and tail checks using skewness, 1st/99th percentiles and a top-10 up vs top-10 down comparison (with earnings days flagged). The edge is real but modest: a few outsized events, including earnings-linked moves, do a lot of the work. Full numbers, charts and the head-to-head breakdown follow below.
For PLTR over the past ~3 years, does the 'stocks take the stairs up and the elevator down' cliché actually hold — are its largest single-day moves skewed to the downside, or does this momentum darling deliver its biggest lurches to the UPSIDE on breakouts and earnings? Thesis: PLTR's daily-return distribution is positively skewed with its fattest tail pointing up, so its biggest one-day shocks have been melt-ups rather than crashes, flipping the folk wisdom on its head.
How this was measured
Resampled PLTR minute bars to daily closes and computed close-to-close returns over the trailing 36 months (bounded by the available data range). Assessed tail asymmetry via Fisher-Pearson skewness, extreme quantiles (1st/99th percentiles), and a head-to-head comparison of the top-10 positive vs top-10 negative one-day moves by magnitude. Ranked the largest absolute daily moves and flagged whether each fell on the first trading day on/after an earnings reported_date (earnings often catalyze outsized moves).
The key numbers
Reading the numbers
Across 751 trading days PLTR shows a small positive bias and modest positive skew (skewness ≈0.420). Its single biggest day was a +24.51% up move versus a largest down day of -15.08%, and 55% of the top-20 absolute moves were upward.
The charts
The histogram piles most daily returns close to zero with a slight tilt to the right — the mean daily return is about 0.35% and the median is about 0.347%. Look at the center cluster (many small moves) and then the longer right tail stretching to the max near +24.51% vs the left extreme at -15.08%; given a daily std of ~4.07%, those extremes are much larger than a typical day. In plain terms, routine days are small, but the distribution has a fatter positive tail, consistent with more extreme up days than down days.
This bar chart lines up the top-10 absolute moves and shows the UP bars exceed the DOWN bars at every rank: the largest up is 24.51% versus the largest down at 15.08%, and at ranks 2–10 the upside magnitudes (e.g., 21.48%, 16.54%, 14.02%) are consistently higher than their downside counterparts. The mean of the top-10 UP moves is about 14.88% versus about 11.92% for the top-10 DOWN moves, so the biggest shocks have been larger on the upside. That pattern directly supports flipping the 'stairs up, elevator down' cliché for this sample period: PLTR's biggest one-day lurches have tended to be melt-ups, not crashes.
Top 15 absolute one-day moves (earnings-flagged)
| date | return | direction | abs_rank | earnings_day |
|---|---|---|---|---|
| 2025-02-03 | 0.2451 | Up | 1 | Yes |
| 2025-04-09 | 0.2148 | Up | 2 | No |
| 2023-11-02 | 0.1654 | Up | 3 | Yes |
| 2025-03-10 | -0.1508 | Down | 4 | No |
| 2025-02-19 | -0.1491 | Down | 5 | No |
| 2024-08-08 | 0.1402 | Up | 6 | No |
| 2024-02-05 | 0.1375 | Up | 7 | Yes |
| 2023-07-28 | 0.135 | Up | 8 | No |
| 2025-04-04 | -0.1344 | Down | 9 | No |
| 2025-02-24 | -0.1243 | Down | 10 | No |
| 2024-11-04 | 0.1236 | Up | 11 | Yes |
| 2025-02-06 | 0.1185 | Up | 12 | No |
| 2023-08-24 | -0.1176 | Down | 13 | No |
| 2025-08-19 | -0.1132 | Down | 14 | No |
| 2024-02-06 | 0.105 | Up | 15 | No |
The takeaway
Yes — over the trailing ~36 months PLTR's biggest single-day moves lean to the upside. The return series shows a positive skew of about +0.42, the largest up day was +24.51% (2025-02-03) while the largest down day was -15.08% (2025-03-10). Looking at extremes, 55% of the top-20 absolute moves were up, the mean of the top-10 up days was 14.88% versus 11.92% for the top-10 down days (a tail-edge of +2.96%), and of 124 days with |return| ≥ 5% about 58% closed higher. This is a real, modest tilt rather than a slam-dunk: with 751 trading days the signal is suggestive but narrow, and a few outsized events (including earnings-linked moves such as 2025-02-03) do a lot of the heavy lifting. Practical takeaway: don’t assume PLTR’s shocks are mainly crashes — over this window melt-ups dominate slightly, but the margin is small and sensitive to a handful of extreme days.
The fine print
- Window limited to the trailing ~36 months; earlier regimes (IPO/initial volatility) are excluded.
- Earnings-day flag uses the first trading day on/after reported_date; post-market releases can move the reaction to the next day.
- Returns are close-to-close; intraday spikes or gap+reversal patterns are compressed into single daily changes.
- Tail metrics are driven by a few outliers; a different window or one extreme event could flip the tilt.