Skip to content
FrontierPicks

· an extra edition, outside the schedule

Tested patterns · re-tested each quarter

What the market data shows: tested patterns in US stocks.

Tested as of , a run outside the quarterly schedule (the first Saturday of January, April, July and October)

Beliefs about stock prices are often repeated and rarely checked: that a stock which jumped keeps rising, that the most-traded stocks beat the market, that cash is safer when the market is below its trend. Each is tested here on the daily prices of liquid US stocks since 2017, period by period, and tested again every quarter.

Prices through

The price history holds 10,378 stocks, 4,914 of which stopped trading. It holds 97.4% of the exchange-listed common stocks since Jan 2018; a verdict is given only when it would stand even if the missing ones moved the result by 1.43 points, as if each had ended in a delisting loss of 55%.

None of the 71 tested lines holds.

  • Ruled out: After a one-day jump, $5-20M a day, next 63 sessions: the claimed −8.5% lies outside the Nov 2023–Oct 2026 range, −2.5% to +1.5%.
  • Too little data to tell: 70 lines, the base rates by situation included. A line that does not hold is shown with the range its effect could lie in: too little data to tell, unless that range rules out the size a stated belief claims.
  • 1 description of the market, not a test, shown for what it tells.
How every test is run, and when a line holds
  • A line holds when both deciding periods are at least 2.43 standard errors from zero with the same sign and the checks agree (slightly more where a period has few independent days of data). That bar is set so that, across the 445 tests counted (71 lines here, 4 registered studies and 370 tried earlier on the same prices), chance alone would produce no more than 0.05 holding lines a run on average.
  • Each line is measured in periods of 36 months; the two latest, Nov 2020–Oct 2023 and Nov 2023–Oct 2026, decide.
  • One event per stock per 63 sessions; events averaged by calendar month, months weighted equally. The uncertainty of each average comes from the daily returns inside every holding window, because overlapping windows are not independent and days nearly are.
  • Every situation is defined on information from before the event, and the sign must survive entry at the next close, a same-tier benchmark, and removing names that stopped trading or traded flat.

Beliefs registered before testing

Each study was written down before its results were seen, so a first result that happens to look good cannot be picked afterwards.

Fresh-results screen, forward test

Companies whose filing passes the fresh-results screen beat the S&P 500 over the next 63 sessions by the +3.0% the in-sample history claims.

The belief: +3.0% over the next 63 sessions against the S&P 500. Where it is stated: this project's own back-test of 2016 to 2026, in-sample and unconfirmed.

Not run yet · It is tested only on filings from 16 November 2026 on, and none of them has had its 63 trading days yet.

Registered on 14 November 2026; tested only on prices after that day. Amended on 14 November 2026: Before any run, the precision rule was made stricter: a verdict is given only once the test would find an effect of the claimed size 80% of the time.

Share dilution

Companies whose diluted share count rose most over a year trail the average stock of their sector and tier over the next 63 sessions.

The belief: −1.0% over the next 63 sessions against the other stocks of its sector and trading-volume tier. Where it is stated: Fama and French (2008), 'Dissecting Anomalies', Journal of Finance 63(4), Table II.

Too little data to tell, because its range is still too wide to tell · Nov 2020–Oct 2023: −2.1% (−6.6 to +2.3), 1,974 events · Nov 2023–Oct 2026: −3.1% (−9.0 to +2.7), 2,329 events

Registered on 14 November 2026; tested on prices from 4 January 2016. Amended on 14 November 2026: Before any run: the signal became each filing's own comparison of its diluted share count with the same period a year earlier, because companies do not file a fourth quarter; each stock is measured against the other stocks of its group, not a group that includes itself; and a verdict is given only once the test would find an effect of the claimed size 80% of the time.

The 126-session figures, for information

For information only, it decides nothing: over 126 sessions, Jan 2017–Oct 2017: −0.8% (−5.6 to +4.1) · Nov 2017–Oct 2020: +10.3% (−1.3 to +21.8) · Nov 2020–Oct 2023: −6.3% (−12.9 to +0.4) · Nov 2023–Oct 2026: −4.1% (−14.2 to +6.0).

Gross profitability

Companies with the highest gross profit relative to their assets beat the other stocks of their sector and tier over the next 63 sessions.

The belief: +0.5% over the next 63 sessions against the other stocks of its sector and trading-volume tier. Where it is stated: Novy-Marx (2013), 'The Other Side of Value: The Gross Profitability Premium', Journal of Financial Economics 108(1), Table 2.

Too little data to tell, because its range is still too wide to tell · Nov 2020–Oct 2023: +1.6% (−1.7 to +5.0), 364 events · Nov 2023–Oct 2026: −4.7% (−8.8 to −0.6), 348 events

Registered on 15 November 2026; tested on prices from 4 January 2016.

The 126-session figures, for information

For information only, it decides nothing: over 126 sessions, Jan 2017–Oct 2017: −6.9% (−17.0 to +3.2) · Nov 2017–Oct 2020: +1.5% (−10.2 to +13.2) · Nov 2020–Oct 2023: +0.9% (−4.0 to +5.8) · Nov 2023–Oct 2026: −7.9% (−15.3 to −0.5).

Insider purchases

Stocks whose officers or directors bought at least $100,000 on the open market within five sessions beat the average stock of their sector and tier over the next 63 sessions.

The belief: +1.1% over the next 63 sessions against the other stocks of its sector and trading-volume tier. Where it is stated: Jeng, Metrick and Zeckhauser, 'Estimating the Returns to Insider Trading', Review of Economics and Statistics 85(2), 2003 (NBER Working Paper 6913, Table 5).

Too little data to tell, because its range is still too wide to tell · Nov 2020–Oct 2023: +0.3% (−1.2 to +1.6), 2,196 events · Nov 2023–Oct 2026: +0.5% (−0.9 to +1.9), 2,157 events

Registered on 14 November 2026; tested on prices from 4 January 2016. Amended on 14 November 2026: Before any run: the reading of the rule was written down (which filings and purchases count, and that a purchase needs a price that matches that day's trading); each stock is measured against the other stocks of its group, not a group that includes itself; and a verdict is given only once the test would find an effect of the claimed size 80% of the time.

The 126-session figures, for information

For information only, it decides nothing: over 126 sessions, Jan 2017–Oct 2017: −0.1% (−2.8 to +2.5) · Nov 2017–Oct 2020: +1.2% (−1.3 to +3.8) · Nov 2020–Oct 2023: −0.5% (−2.6 to +1.6) · Nov 2023–Oct 2026: +0.2% (−2.2 to +2.5).

Short interest

Stocks with the most days to cover trail the average stock of their tier over the next 63 sessions.

The belief: −0.8% over the next 63 sessions against the other stocks of its trading-volume tier. Where it is stated: Hong, Li, Ni, Scheinkman and Yan (2015), 'Days to Cover and Stock Returns', NBER Working Paper 21166, Tables 3 and 4.

Too little data to tell, because its range is still too wide to tell · Nov 2020–Oct 2023: −2.2% (−3.8 to −0.7), 4,029 events · Nov 2023–Oct 2026: +0.6% (−0.8 to +2.0), 4,000 events

Registered on 14 November 2026; tested on prices from 4 January 2016. Amended on 14 November 2026: Before any run: the reading of the rule was written down (how the most-shorted tenth is chosen, and that a figure counts only from the day it was published); each stock is measured against the other stocks of its group, not a group that includes itself; and a verdict is given only once the test would find an effect of the claimed size 80% of the time.

The 126-session figures, for information

For information only, it decides nothing: over 126 sessions, Nov 2017–Oct 2020: +0.5% (−2.5 to +3.4) · Nov 2020–Oct 2023: −5.0% (−7.6 to −2.4) · Nov 2023–Oct 2026: +0.3% (−2.5 to +3.1).

The average stock against the S&P 500

5 tests, each with too little data to tell
The average stock against the S&P 500: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
Average liquid stock against the S&P 500 Nov 2020–Oct 2023: −7.7 (−17.6 to +2.1) · Nov 2023–Oct 2026: −4.2 (−13.5 to +5.1) points a year
Average $5-20M stock against the S&P 500 Nov 2020–Oct 2023: −10.2 (−22.6 to +2.2) · Nov 2023–Oct 2026: −4.5 (−16.6 to +7.7) points a year
Average $20-100M stock against the S&P 500 Nov 2020–Oct 2023: −7.0 (−16.9 to +2.8) · Nov 2023–Oct 2026: −5.1 (−14.8 to +4.6) points a year
Average $100M-1B stock against the S&P 500 Nov 2020–Oct 2023: −4.8 (−11.8 to +2.1) · Nov 2023–Oct 2026: −3.1 (−10.3 to +4.1) points a year
Average over $1B stock against the S&P 500 Nov 2020–Oct 2023: −5.3 (−15.4 to +4.8) · Nov 2023–Oct 2026: +6.0 (−3.9 to +15.9) points a year

How far apart stocks end up

Spread of 63-session returns across liquid stocks

A description, not a test · 2025: 26.4% · 2026: 26.8% (standard deviation)

What was tested, every period and the method

The standard deviation of 63-session returns across all liquid stocks, by year: how far apart the best and worst stocks end up, and so how much picking the right ones can be worth.

Spread of 63-session returns across liquid stocks: result by period
Period Average, % (standard deviation) 90% range Median Events Independent windows
2017 16.4 16.3 251 —
2018 17.9 17.8 251 —
2019 18.6 18.4 252 —
2020 32.6 30.0 253 —
2021 21.5 20.8 252 —
2022 21.2 20.9 251 —
2023 21.3 20.9 250 —
2024 22.3 21.4 252 —
2025 · decides 26.4 27.0 250 —
2026 · decides 26.8 26.0 156 —
Jan 2017–Dec 2026 (all) 22.4 20.9 2,418 —

Every session, the cross-sectional standard deviation of the next 63 sessions' returns (capped at +1000%, stopped stocks at their last close); sessions averaged by calendar month. A description, not a test: a spread is never zero, so it has no state.

After a one-day jump

After a one-day jump, $5-20M a day, next 63 sessions

Rules out the belief · Nov 2020–Oct 2023: −1.1 (−2.8 to +0.5)% · Nov 2023–Oct 2026: −0.5 (−2.5 to +1.5)%

The belief: the claimed −8.5% lies outside the Nov 2023–Oct 2026 range, −2.5% to +1.5%. Where it is stated: DESIGN-V9.md section 2: 'Thin stocks fade 8.5% after a news jump' (withdrawn as an artefact; research/v9-review/tests_verify.py (1) tested it at 63 sessions against the average stock).

What was tested, every period and the method

Stocks liquid the day before ($5-20M a day) with a one-day move of at least 3 standard deviations of its prior 63 sessions and at least 4% against the S&P 500 (up): their return over the next 63 sessions against the average liquid stock.

After a one-day jump, $5-20M a day, next 63 sessions: result by period
Period Average, % over 63 sessions 90% range Median Events Independent windows
Jan 2017–Oct 2017 −0.2 −2.0 to +1.6 −0.9 910 4
Nov 2017–Oct 2020 −0.2 −1.6 to +1.3 −2.1 3,683 13
Nov 2020–Oct 2023 · decides −1.1 −2.8 to +0.5 −2.0 4,007 13
Nov 2023–Oct 2026 · decides −0.5 −2.5 to +1.5 −3.3 3,679 12
Jan 2017–Oct 2026 (all) −0.6 −1.5 to +0.3 −2.2 12,279 39

Measured against: average liquid stock.

From Nov 2020–Oct 2023 to Nov 2023–Oct 2026 the average moved by +0.7 (−1.9 to +3.2).

One event per stock per 63 sessions; events averaged within the calendar month, months weighted equally; the uncertainty from the daily returns of every stock inside its holding window, less the benchmark's. Liquid: median dollar volume over the prior 63 sessions at least $5M, close at least $3 a share and 252 sessions of history, all at the close before the entry. A stock that stops trading keeps its last close; returns capped at +1000%. Checked with entry at the next close, against the average stock of the same liquidity tier, and without names that stopped trading or traded flat (daily moves under 0.5%, as when pinned to a deal price).

In the deciding periods every check keeps the direction of the main result.

After a one-day jump, $5-20M a day, next 63 sessions: the average under each check
Period Main result Bought at the next closeAgainst stocks of the same liquidityWithout stocks that stopped trading or traded flat
Jan 2017–Oct 2017 −0.2 0.0−0.1−0.1
Nov 2017–Oct 2020 −0.2 −0.3+0.1−0.1
Nov 2020–Oct 2023 −1.1 −1.0−0.7−1.2
Nov 2023–Oct 2026 −0.5 −0.3−0.5−0.7
9 more tests with too little data to tell
After a one-day jump: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
After a one-day jump, next 21 sessions Nov 2020–Oct 2023: 0.0 (−0.6 to +0.5)% · Nov 2023–Oct 2026: 0.0 (−0.6 to +0.7)%
After a one-day jump, next 63 sessions Nov 2020–Oct 2023: −1.0 (−1.9 to 0.0)% · Nov 2023–Oct 2026: −0.6 (−1.5 to +0.3)%
After a one-day jump, $5-20M a day, next 21 sessions Nov 2020–Oct 2023: −0.2 (−0.9 to +0.6)% · Nov 2023–Oct 2026: −0.2 (−1.5 to +1.1)%
After a one-day jump, $20-100M a day, next 21 sessions Nov 2020–Oct 2023: −0.2 (−0.8 to +0.4)% · Nov 2023–Oct 2026: +0.1 (−0.6 to +0.7)%
After a one-day jump, $20-100M a day, next 63 sessions Nov 2020–Oct 2023: −1.2 (−2.4 to −0.1)% · Nov 2023–Oct 2026: −0.6 (−1.5 to +0.4)%
After a one-day jump, $100M-1B a day, next 21 sessions Nov 2020–Oct 2023: +1.1 (+0.1 to +2.0)% · Nov 2023–Oct 2026: 0.0 (−0.7 to +0.7)%
After a one-day jump, $100M-1B a day, next 63 sessions Nov 2020–Oct 2023: +0.8 (−2.9 to +4.5)% · Nov 2023–Oct 2026: −0.4 (−1.9 to +1.0)%
After a one-day jump, over $1B a day, next 21 sessions Nov 2020–Oct 2023: 0.0 (−2.1 to +2.1)% · Nov 2023–Oct 2026: +2.0 (−0.2 to +4.2)%
After a one-day jump, over $1B a day, next 63 sessions Nov 2020–Oct 2023: −1.5 (−5.5 to +2.6)% · Nov 2023–Oct 2026: +6.3 (+1.0 to +11.6)%

After a one-day drop

10 tests, each with too little data to tell
After a one-day drop: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
After a one-day drop, next 21 sessions Nov 2020–Oct 2023: +0.1 (−0.5 to +0.6)% · Nov 2023–Oct 2026: +0.6 (+0.2 to +1.0)%
After a one-day drop, next 63 sessions Nov 2020–Oct 2023: −0.6 (−1.5 to +0.3)% · Nov 2023–Oct 2026: +0.3 (−0.5 to +1.1)%
After a one-day drop, $5-20M a day, next 21 sessions Nov 2020–Oct 2023: 0.0 (−0.8 to +0.7)% · Nov 2023–Oct 2026: +1.1 (+0.2 to +1.9)%
After a one-day drop, $5-20M a day, next 63 sessions Nov 2020–Oct 2023: −1.3 (−2.8 to +0.1)% · Nov 2023–Oct 2026: +1.3 (−0.4 to +3.1)%
After a one-day drop, $20-100M a day, next 21 sessions Nov 2020–Oct 2023: +0.3 (−0.4 to +0.9)% · Nov 2023–Oct 2026: +0.2 (−0.3 to +0.8)%
After a one-day drop, $20-100M a day, next 63 sessions Nov 2020–Oct 2023: +0.1 (−1.1 to +1.3)% · Nov 2023–Oct 2026: −0.8 (−1.9 to +0.3)%
After a one-day drop, $100M-1B a day, next 21 sessions Nov 2020–Oct 2023: +0.5 (−0.4 to +1.3)% · Nov 2023–Oct 2026: +0.5 (−0.1 to +1.2)%
After a one-day drop, $100M-1B a day, next 63 sessions Nov 2020–Oct 2023: +0.4 (−1.2 to +2.0)% · Nov 2023–Oct 2026: +0.6 (−0.9 to +2.1)%
After a one-day drop, over $1B a day, next 21 sessions Nov 2020–Oct 2023: −0.8 (−2.9 to +1.4)% · Nov 2023–Oct 2026: +1.0 (−1.3 to +3.3)%
After a one-day drop, over $1B a day, next 63 sessions Nov 2020–Oct 2023: −1.2 (−5.0 to +2.6)% · Nov 2023–Oct 2026: +0.7 (−4.1 to +5.5)%

After a rise, a fall or a pullback

14 tests, each with too little data to tell
After a rise, a fall or a pullback: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
After a rise of 50% from entry, next 63 sessions Nov 2020–Oct 2023: −0.4 (−4.6 to +3.8)% · Nov 2023–Oct 2026: +1.2 (−2.2 to +4.5)%
After a rise of 50% from entry, next 126 sessions Nov 2020–Oct 2023: +0.6 (−6.9 to +8.1)% · Nov 2023–Oct 2026: +3.3 (−3.4 to +10.0)%
After a rise of 100% from entry, next 63 sessions Nov 2020–Oct 2023: +0.7 (−5.5 to +6.8)% · Nov 2023–Oct 2026: +1.9 (−3.5 to +7.2)%
After a rise of 100% from entry, next 126 sessions Nov 2020–Oct 2023: +3.5 (−7.8 to +14.8)% · Nov 2023–Oct 2026: +4.4 (−6.4 to +15.2)%
After a fall of 30% from entry, next 63 sessions Nov 2020–Oct 2023: −0.5 (−3.9 to +3.0)% · Nov 2023–Oct 2026: 0.0 (−3.0 to +2.9)%
After a fall of 30% from entry, next 126 sessions Nov 2020–Oct 2023: −0.8 (−6.2 to +4.5)% · Nov 2023–Oct 2026: +0.9 (−4.3 to +6.2)%
After a fall of 50% from entry, next 63 sessions Nov 2020–Oct 2023: −3.5 (−8.5 to +1.5)% · Nov 2023–Oct 2026: +3.0 (−2.7 to +8.6)%
After a fall of 50% from entry, next 126 sessions Nov 2020–Oct 2023: −8.7 (−16.9 to −0.4)% · Nov 2023–Oct 2026: +6.3 (−3.4 to +15.9)%
After doubling, then 25% below the peak, next 63 sessions Nov 2020–Oct 2023: −5.0 (−12.1 to +2.1)% · Nov 2023–Oct 2026: +5.4 (−3.0 to +13.8)%
After doubling, then 25% below the peak, next 126 sessions Nov 2020–Oct 2023: −3.8 (−15.3 to +7.6)% · Nov 2023–Oct 2026: +8.0 (−6.9 to +22.8)%
After doubling, then 35% below the peak, next 63 sessions Nov 2020–Oct 2023: −1.6 (−10.4 to +7.2)% · Nov 2023–Oct 2026: +9.4 (−1.5 to +20.3)%
After doubling, then 35% below the peak, next 126 sessions Nov 2020–Oct 2023: +2.0 (−12.0 to +16.0)% · Nov 2023–Oct 2026: +12.5 (−6.7 to +31.6)%
After doubling, then 50% below the peak, next 63 sessions Nov 2020–Oct 2023: −0.2 (−12.2 to +11.7)% · Nov 2023–Oct 2026: +4.2 (−6.5 to +15.0)%
After doubling, then 50% below the peak, next 126 sessions Nov 2020–Oct 2023: +2.1 (−20.4 to +24.6)% · Nov 2023–Oct 2026: +3.7 (−14.2 to +21.6)%

Stocks that trade heavily on a few days only

1 test, with too little data to tell
Stocks that trade heavily on a few days only: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
Liquid on the mean but not the median, next 63 sessions Nov 2020–Oct 2023: −7.0 (−12.6 to −1.3)% · Nov 2023–Oct 2026: −13.4 (−17.9 to −9.0)%

The most-traded stocks

3 tests, each with too little data to tell
The most-traded stocks: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
The 30 most-traded stocks against the S&P 500 Nov 2020–Oct 2023: −6.5 (−18.1 to +5.1) · Nov 2023–Oct 2026: +8.5 (−5.6 to +22.6) points a year
The 50 most-traded stocks against the S&P 500 Nov 2020–Oct 2023: −6.8 (−15.9 to +2.3) · Nov 2023–Oct 2026: +6.6 (−4.7 to +17.9) points a year
The 100 most-traded stocks against the S&P 500 Nov 2020–Oct 2023: −6.8 (−14.8 to +1.2) · Nov 2023–Oct 2026: +2.8 (−5.8 to +11.3) points a year

How a portfolio's stocks are weighted

2 tests, each with too little data to tell
How a portfolio's stocks are weighted: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
Sizing by inverse volatility against equal weight Nov 2020–Oct 2023: +1.3 (−1.8 to +4.3) · Nov 2023–Oct 2026: −0.9 (−4.0 to +2.2) points a year
Sizing by inverse variance against equal weight Nov 2020–Oct 2023: +2.1 (−3.0 to +7.1) · Nov 2023–Oct 2026: −1.9 (−7.1 to +3.3) points a year

Holding cash when the market is below its trend

1 test, with too little data to tell
Holding cash when the market is below its trend: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
Cash while the S&P 500 is below its 200-day average Nov 2020–Oct 2023: +1.0 (−9.7 to +11.6) · Nov 2023–Oct 2026: −4.0 (−11.1 to +3.2) points a year

What skill in ranking stocks is worth

1 test, with too little data to tell
What skill in ranking stocks is worth: tests with too little data to tell, with the two deciding periods
TestThe deciding periods, average and 90% range
Ranking skill and breadth for a 20-name book Nov 2020–Oct 2023: −7.5 (−17.4 to +2.4) · Nov 2023–Oct 2026: −4.0 (−13.4 to +5.4) points a year

What an average stock did next, by situation

Formed every session on 5 staggered calendars: each liquid stock of the tier, with at least 252 sessions of prices, is sorted by its last 5 sessions into jump, drop or neither, bought at that close and held 21 or 63 sessions, against the average liquid stock bought the same close. Each cell is the average return against the average liquid stock in the latest period (Nov 2023–Oct 2026), with its 90% range. For every one of them there is too little data to tell.

What each row means
$100M-1B:
Median dollar volume over the last 63 sessions of $100 million to $1 billion a day, read 5 sessions before the day measured (so a recent move never selects the stock), at a price of at least $3 a share.
$20-100M:
Median dollar volume over the last 63 sessions of $20 million to $100 million a day, read 5 sessions before the day measured (so a recent move never selects the stock), at a price of at least $3 a share.
$5-20M:
Median dollar volume over the last 63 sessions of $5 million to $20 million a day, read 5 sessions before the day measured (so a recent move never selects the stock), at a price of at least $3 a share.
over $1B:
Median dollar volume over the last 63 sessions of $1 billion a day or more, read 5 sessions before the day measured (so a recent move never selects the stock), at a price of at least $3 a share.
Drop:
The stock's most recent such move in the last 5 sessions was a fall: a one-day fall of at least 4% against the S&P 500 that is also at least 3 times the stock's usual daily move against it (the standard deviation of the 63 sessions before, at least 31 of them on file).
Jump:
The stock's most recent such move in the last 5 sessions was a rise: a one-day rise of at least 4% against the S&P 500 that is also at least 3 times the stock's usual daily move against it (the standard deviation of the 63 sessions before, at least 31 of them on file).
Neither:
No one-day rise or fall of that size in the last 5 sessions.
Base rates by trading volume and recent move
Trades a day, and recent move Next 21 sessions, %Next 63 sessions, %
$5-20MJump −0.3 −1.5 to +0.8 −0.5 −2.3 to +1.3
$5-20MDrop +1.0 +0.3 to +1.7 +1.4 −0.3 to +3.1
$5-20MNeither −0.3 −0.7 to +0.1 0.0 −1.0 to +1.0
$20-100MJump 0.0 −0.7 to +0.6 −0.8 −1.7 to +0.2
$20-100MDrop 0.0 −0.6 to +0.5 −0.8 −1.9 to +0.2
$20-100MNeither −0.1 −0.2 to +0.1 −0.2 −0.5 to +0.2
$100M-1BJump 0.0 −0.7 to +0.7 −0.3 −1.7 to +1.2
$100M-1BDrop +0.2 −0.5 to +0.8 +0.6 −0.8 to +2.1
$100M-1BNeither −0.1 −0.5 to +0.3 +0.2 −0.9 to +1.2
over $1BJump +1.8 −0.3 to +3.9 +5.5 +0.3 to +10.8
over $1BDrop +0.3 −1.7 to +2.3 +1.8 −2.7 to +6.2
over $1BNeither +0.1 −1.1 to +1.3 +2.5 −0.8 to +5.8