American Express shareholders face a $350 million regulatory penalty against its bank subsidiary. How much comfort can the shares’ seasonal history offer while the earnings implications remain unestablished?[1][3]
Ahead of October 24–November 10, 2026, TradeWave finds gains in all ten selected midterm-election years spanning 1986–2022. That favorable seasonal record warrants attention, although comparisons across all years include losses.[3]
Growth and compliance failures
The OCC’s October 8 cease-and-desist order cited failures in staffing, controls, testing, training, risk assessment and customer checks. Approximately $13 billion concerned suspected trade-based money laundering over the past decade that the bank failed to identify, evaluate and sufficiently report promptly. It was not the penalty amount.[1]
July 24’s second-quarter release reported revenue net of interest expense of $19.6 billion, up 10% year over year, and diluted EPS of $4.53 versus $4.08 in Q2 2025. Management credited spending, net interest income supported by card-balance growth, and card fees. Expenses grew 12%, faster than revenue. Citing first-half outperformance, management raised its 2026 revenue-growth forecast to 10%, planned reinvestment in growth initiatives and maintained its $17.30–$17.90 EPS forecast.[2]
TradeWave estimates earnings on October 23; the saved primary sources do not confirm a release date.[3]
American Express second-quarter diluted EPS
Reported earnings per common share
View chart data & download
| Measure | Value | Period | Status |
|---|---|---|---|
| Q2 2025 | 4.08 | Q2 2025 | reported |
| Q2 2026 | 4.53 | Q2 2026 | reported |
What the selected windows show
TradeWave’s selected midterm-year study puts the median full-window long result at 6.2%. This provides historical timing context, but ten observations cannot establish an election-driven advantage or predict this year’s return.[3]
The year-by-year seasonal record
View exact TradeWave observations and download
| Window starting year | Ending price change | Highest from entry | Lowest from entry |
|---|---|---|---|
| 1986 | +4.18% | +5.49% | +0.00% |
| 1990 | +6.79% | +7.40% | -12.96% |
| 1994 | +4.24% | +6.36% | -0.42% |
| 1998 | +5.60% | +11.13% | -3.47% |
| 2002 | +7.95% | +20.12% | -3.10% |
| 2006 | +3.81% | +4.94% | +0.00% |
| 2010 | +10.84% | +12.65% | -0.41% |
| 2014 | +6.94% | +7.04% | -0.74% |
| 2018 | +4.57% | +7.35% | -1.77% |
| 2022 | +9.26% | +10.00% | -2.02% |
Explore this exact seasonal window in TradeWave. Compare its win rate and average result across different histories, and inspect how far prices moved above or below their starting point. Account access applies in TradeWave.
A gain can conceal discomfort
The yearly range chart covers the same seasonal period each year, not whole calendar years. October 24–November 10, 1990 is useful because its 6.79% ending gain coexisted with a move 12.96% below the starting price. Those highs and lows reveal exposure hidden by the ending result, without establishing their order or a peak-to-trough loss.[3]
What happened inside each seasonal window
View exact TradeWave observations and download
| Window starting year | Ending price change | Highest from entry | Lowest from entry |
|---|---|---|---|
| 1986 | +4.18% | +5.49% | +0.00% |
| 1990 | +6.79% | +7.40% | -12.96% |
| 1994 | +4.24% | +6.36% | -0.42% |
| 1998 | +5.60% | +11.13% | -3.47% |
| 2002 | +7.95% | +20.12% | -3.10% |
| 2006 | +3.81% | +4.94% | +0.00% |
| 2010 | +10.84% | +12.65% | -0.41% |
| 2014 | +6.94% | +7.04% | -0.74% |
| 2018 | +4.57% | +7.35% | -1.77% |
| 2022 | +9.26% | +10.00% | -2.02% |
Broader histories temper confidence
Checking across all years tests how dependent the finding is on selected midterms: profitable windows fall from 80.0% in 2016–2025 to 70.0% in 2006–2025. The favorable reading becomes less consistent further back. Both comparisons reuse midterm observations, and the longer includes the entire shorter sample, limiting independent confirmation and any claim of an election effect.[3]
See the history behind this comparison
TradeWave checks the same seasonal dates across different sets of years. This helps show whether the pattern holds up when you look at another stretch of history.
| History | Observed start years | Direction | Success rate | Median result |
|---|---|---|---|---|
| Article study: 10 selected midterm-election years (1986–2022) | 1986, 1990, 1994, 1998, 2002, 2006, 2010, 2014, 2018, 2022 | long | 100.0% | 6.2% |
| 10 consecutive historical windows (2016–2025) | 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025 | long | 80.0% | 4.54% |
| 20 consecutive historical windows (2006–2025) | 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025 | long | 70.0% | 3.84% |
Some years appear in more than one group, so agreement is not an independent test. A requested lookback can extend beyond available history; the observed years are listed above. Election-cycle groups use matching phases, not consecutive years. Results follow the direction shown: a positive short result means prices fell. Small samples and different market periods make the comparison less conclusive.
Inspect the path and next disclosure
Next, seek confirmed earnings timing and disclosed penalty implications. Below, recorded prices end October 20, 2026, before the future window entry. The engine’s seasonal overlay uses this article’s ten selected midterm years over the next 60 weekdays, October 21, 2026–January 12, 2027. That separate horizon illustrates historical shape, not future recorded prices or a target. In TradeWave, compare this same window across histories and inspect its range of outcomes.[3][1]
Price history and the seasonal outlook
Explore this exact seasonal window in TradeWave. Compare its win rate and average result across different histories, and inspect how far prices moved above or below their starting point. Account access applies in TradeWave.
About this seasonal analysis
10 selected midterm-election years (1986–2022). The selected dates, years, direction, returns and ranges come from TradeWave. When a window boundary is a weekend or market holiday, TradeWave uses the next available trading session. SMN displays the returned results without recalculating them. A year label identifies the start of the historical window; a window can end in the following calendar year.
Green and red bars show the underlying adjusted-price movement. The range endpoints measure changes from the entry price, not a peak-to-trough loss. Historical results are before trading costs and do not guarantee future performance.
Sources & methodology
- OCC Assesses $350 Million Civil Money Penalty Against American Express · 2026-10-08
- AMERICAN EXPRESS Q2 2026 RESULTS · 2026-07-24
- TradeWave AXP seasonal evidence · 2026-10-21