How to Use Claude for Earnings Season: 5 Questions to Ask
How to use Claude for earnings season with the SentiSense connector: who reports this week, why a stock fell after its report, which number it beat on, how its options are positioned, and what last week's bank earnings say about the next. With the connector's own output from October 16, 2026.
Earnings season is when a market question comes with a date on it. With the SentiSense connector added, Claude answers five of those questions from dated data: who reports this week and when, why a stock moved after its report, whether it beat and on which number, how its options are positioned before the report, and what one sector's results say about the companies reporting next.
SentiSense is listed in Claude's connector directory; the setup is in Stock Market Data in Claude. Everything quoted below is the connector's own output on Friday, October 16, 2026, the week the big banks reported.
Who reports earnings this week?
Ask "who reports earnings next week, before or after the bell?" The calendar tool returns each company's date, timing, whether the date is confirmed and the consensus EPS. On October 16 it returned (condensed):
Earnings next week: Mon 2026-10-19 to Sun 2026-10-25. 137 companies report next week, 117 confirmed. NFLX Tue AMC. META (consensus EPS $6.74), TSLA ($0.46), IBM, TXN, LRCX Wed AMC. PG, HON, LMT, BX Thu BMO.
AMC means after the close, so the report trades the next session. Every tier sees the requested week in full; the ranked view shows the top three reports per section on a free account and twelve on PRO.
Why did the stock drop after earnings?
The most-searched question of the season needs three reads: the numbers against consensus, the guidance, and what the market did that day. Bank of America, from the ranked earnings view on October 16:
BAC, reported 2026-10-14. EPS surprise -1.82%, reaction -4.76%. $0.91 diluted EPS after a $1.2B regulatory settlement charge; adjusted $1.08; revenue up 3% to $30.0B.
Ask "why did BAC fall after earnings?" and Claude adds the news tool: a six-source story cluster, sentiment -0.67, traced the EPS to a $0.17-a-share charge for the Treasury penalty announced October 13. Then the tape: SPY fell 1.08% that session and the financials ETF XLF 0.55% (SentiSense daily closes), so part of BAC's move was the market. Compare the move with the market and the sector that day before blaming the report.
Guidance is often the swing factor, and the connector has no guidance field: it reaches guidance through the news stories and the ranked view's one-line headline.
Did it beat, and on which number?
The consensus EPS in the calendar does not say whether it is GAAP or adjusted, and companies report both. BAC's -1.82% surprise is $1.08 against $1.10, so it compares the adjusted figure; against GAAP $0.91 the gap is far wider. JPMorgan reported $5.57, or $5.75 excluding a significant item, against a $5.93 consensus: a miss on both bases. Ask Claude for both figures from the release (the stories carry them) and read the surprise against the matching one.
How are the options positioned before the report?
META reports Wednesday, October 21 after the close, confirmed, consensus EPS $6.74, so its reacting session is Thursday, October 22. The options tool, session Friday, October 16:
META. IV rank 63 (ATM IV 45%). Put/call volume 0.81, 90th percentile. 25-delta skew +2.0 IV points, 75th percentile. Options sentiment -0.42. Expected move next session +/-4.2%.
Two traps. The expected move covers Monday's session, scaled from about 30 days of implied volatility; it is not the earnings move (META's average move across its last eight reports was 7.4% either way, per the SentiSense API on October 16, a history the connector does not serve yet). The implied move for this week's reporters sets that history against what the options market is pricing. And October 16 was the monthly options expiry on a day META fell 3.7%, so the put/call reading is partly expiration-day activity: consistent with hedging, not a bet. An IV rank of 63 before a report is ordinary; implied volatility tends to rise into the print and fall after it. IV rank, put/call and options sentiment are free; skew is PRO.
What do last week's bank earnings say about this week's?
The six largest US banks reported October 13 and 14, and each closed lower on its report day: C -5.6%, BAC -4.8%, WFC -3.1%, MS -2.8%, JPM -0.9%, GS -0.2% (SentiSense daily closes; SPY was flat on October 13 and down 1.08% on October 14). Regional banks and card lenders report next: FITB Monday; KEY, SYF and COF Tuesday; NTRS Wednesday; HBAN Thursday; AXP Friday.
A bank's quarter is three reads. Earnings in context: JPMorgan's $5.57 against $5.07 in the same quarter a year ago, and against the $5.93 consensus. Credit: the provision is what the bank sets aside for losses it expects; net charge-offs are losses already taken. Expectations: a better quarter can still send a stock down if it misses what investors expected. The financials tool serves EPS history on every tier and, on PRO, the bank KPIs: net interest income, provisions, net interest margin and more.
Try it
Add the SentiSense connector to Claude, ask who reports this week, and run the three reads on the first name that moves.
Connect SentiSense in Claude →
Connector output is as SentiSense returned it on October 16, 2026, from a PRO account; the free tier's depth is noted where it differs. Options data is end of day and the connector covers the most-watched US stocks. This article is educational and is not investment advice.