The release sets the open. The call sets the close.
How I'd run an earnings day, and how I'd score it afterward.
The week the largest U.S. banks reported third-quarter 2026 results, four of the six beat analysts' earnings estimates. Most of them still closed lower on the day they reported.
Citigroup beat and fell more than 5% after guiding fourth-quarter markets revenue down about 20%. Wells Fargo beat on earnings, missed on revenue and fell about 3%. Bank of America fell nearly 5% on a trading shortfall and an anti-money-laundering settlement. The market had priced the beat before the release came out. What it hadn't priced was everything else.
That's the first rule of earnings day: the beat is expected. What moves the stock is the story around the numbers.
Three windows, scored separately
An earnings day produces three distinct reactions, and each answers a different question. I measure each against a regional bank index, the SPDR S&P Regional Banking ETF (KRE), so that a sector-wide move isn't mistaken for a verdict on the company.
| Window | Measured from | What it tells you |
|---|---|---|
| The release | Prior close to the opening bell | How the numbers and the press release landed, before anyone heard management |
| The call | Open to close | Whether the prepared remarks and Q&A added confidence or took it away |
| The next day | Close to close | Whether the message held once people had time to think |
Most regional banks release results before the market opens and hold the call that morning. The release trades for hours in the pre-market with no one there to explain it. Those hours are why the release deserves as much care as the call.
The release: answer the obvious question in writing
Every quarter has one number that will draw the first question: expenses up more than expected, a capital ratio stepping down, a guidance range moving. If the release doesn't explain it, the pre-market will price the worst version. A two-line bridge (what drove it, what's one-time, what the run-rate is) is the cheapest insurance in investor relations.
The call: every sentence is a question
Prepared remarks set the Q&A agenda. Analysts follow up on whatever management chooses to emphasize, sometimes a dozen of them on the same point. Before a call I'd test every forward-looking sentence the same way: what question will this trigger, how many analysts will ask it, and how will the stock take the answer?
The best calls hand analysts something new to measure. The worst introduce an uncertainty and leave it open. One unplanned sentence about strategy can undo a strong quarter within the hour.
Earnings day, hour by hour
- Read the largest banks' calls. Their themes reach regional bank calls within days, often through the same analysts.
- Finish the Q&A book. The likely questions, the likely askers, and a one-sentence answer to each. If the release carries strategic news, add a branch for it: why now, what it costs, how fast it pays back, what it does to capital, and what comes next.
- Release and slides post. Watch the pre-market and the first analyst notes for the question nobody planned for.
- The call opens with the market. Track which questions come and how the stock moves with each answer.
- Callbacks begin. Analysts and investors, one after another. Same message and same numbers all day.
- Score the day. Measure the release and the call against the index.
- Score again, in writing. What we predicted, what was asked, how the stock moved, and what we missed.
Score yourself honestly
The scorecard matters most when the preparation was wrong. Topics are easier to predict than askers. In a quiet quarter analysts stay in their lanes, and in a news quarter they abandon them. Writing that down after every call, plainly and without spin, is how the next call's preparation gets better. It's the habit I'd want any team I lead to have.