CommonQuant
CommonQuant.ai Research
AI-generated trading idea · BULLISH · BK, JPM, MS

Banks are the safe corner of a tech crash — long BNY and quality lenders riding earnings beats

The AI trade is unwinding and money is hunting for safe places to sit, which has pushed investors into bonds. Against that backdrop, BNY just showed banks can beat expectations and still rally hard even on a weak tape — its stock went from flat to up more than 3%. If the rest of the bank reporting season confirms this strength, quality bank stocks can act as a defensive haven that also pays a dividend, unlike falling tech names. Elevated rates keep bank lending margins fat, so strong earnings may keep outperforming for the rest of the season.

Idea

The AI trade is unwinding and money is hunting for safe places to sit, which has pushed investors into bonds. Against that backdrop, BNY just showed banks can beat expectations and still rally hard even on a weak tape — its stock went from flat to up more than 3%. If the rest of the bank reporting season confirms this strength, quality bank stocks can act as a defensive haven that also pays a dividend, unlike falling tech names. Elevated rates keep bank lending margins fat, so strong earnings may keep outperforming for the rest of the season.

Advanced Analysis — institutional-depth research report

Verdict: A reasonable bank-haven thesis, but no signal yet — wait for confirmation

The thesis — quality banks as a dividend-paying haven while the AI trade unwinds — has real fundamental support at JPM: Q2 net income of $21.2B was up 28.3% quarter over quarter with return on equity improving from 4.5% to 5.6%, and the trailing dividend of $6.15 per share has grown 10.8% a year, with the latest payment raised to $1.65 (ex-date October 6, 2026). The strongest argument against is that both stocks sit below their 20-day averages today, and the provisional ownership filings for the period ended September 30, 2026 show net insider open-market selling at both names (about $6.6 million at JPM, $8.9 million at MS), so the people closest to the numbers are selling into the strength. This is a watch-list setup, not an active signal: the rules were evaluated on real daily bars but never fired an entry, so the discipline is to wait for confirmation rather than buy anticipation. JPM needs a daily close above $340 (from $334.06) while MS remains far from its $191.95 trigger, and BK could not be evaluated at all because the price feed returned no candles. Because no robust parameter setup was established — the sensitivity evaluation ran out of time before testing any variants — the published 5% stop and 10% target should be treated as design choices, not derived edges. The verdict flips if either stock closes through its trigger with fundamentals intact, or breaks its support ($333.11 for JPM, $178.93 for MS), which would say the pullback is deepening.

Conviction score breakdownComposite score computed by the server from the applicable evidence-tier dimensions.
MeasureValue
Thesis support65/100
Trade readiness40/100
Risk quality45/100
Trigger proximity40/100
Fundamentals trend60/100
Score50/100
Composite Score50/100
Evidence Tierrules_not_triggered
Decision scenariosBull, base, and bear cases synthesized from the cited evidence tier. Likelihoods are rounded evidence-weighted judgments, not statistically calibrated forecasts.
MeasureValue
Evidence Tierrules_not_triggered

Trade now

**Nothing is on yet — and that is the correct reading of today's tape.** This is a watch-list setup: the rules were evaluated on real daily bars but did not open an entry. The trade is a long on two liquid banks (MS first, JPM second), each triggered when four conditions line up at once: a daily close above the 20-day moving average, ADX (14) above 20, RSI (14) below 70, and a close crossing above the nearest resistance level. **JPM is the closer of the two.** At $334.06, the stock is $2.18 below its 20-day average at $336.24 — the price-above-average condition is near but not met. The strength filter is already satisfied (ADX 25.0 vs. a 20 threshold) and momentum is comfortably non-extended (RSI 46.8 vs. a 70 ceiling). The last piece is the breakout: a close above $340, the nearest resistance. So concretely, "wait" for JPM means a daily close above $340 — which would also put the stock through its 20-day average. No robust parameter setup was established, so these published thresholds are the ones to watch, not a variant. **MS is further from firing.** At $179.35 it needs a close above $191.95 — a $12.60 gap to its 20-day average — before the price conditions are even in range. ADX (40.6) and RSI (17.9) are already well inside their thresholds, so MS is a deep-pullback name, not a near-term breakout candidate. BK could not be evaluated: the price feed returned no candles for it, so no live trigger distance can be quoted there. **If an entry fires, the risk envelope is fixed.** Stops are set at a 5% loss on entry price (with a support-based stop also live), targets at +10% or the nearest resistance — for JPM that first target is $340, roughly 1.8% above today's close, so a fresh entry would likely be managed by the trailing rules rather than a wide target. Position sizing is fixed-risk at roughly 2.5% of equity per trade, capped at 25% of the portfolio per name. If neither stock closes above its levels, the disciplined action today is simply to do nothing and let the alert levels come to you.

JPM price and trigger mapUses the idea timeframe and keeps price levels on the price axis.
MeasureValue
TickerJPM
Timeframe1d
MS price and trigger mapUses the idea timeframe and keeps price levels on the price axis.
MeasureValue
TickerMS
Timeframe1d

Bank Earnings Muscle Is Real: Record Profits, Rising Dividends, Shrinking Share Counts

The thesis bets that quality bank stocks act as a defensive haven that pays a dividend while the AI trade unwinds. The reported numbers behind JPM and MS genuinely support the 'strong earnings' leg of that argument. JPMorgan's net income for the quarter ended June 30, 2026 came in at $21.2B, up 28.3% from $16.5B in the prior quarter, with quarterly return on equity improving from 4.53% to 5.65%. Morgan Stanley earned $5.58B, essentially flat versus Q1's $5.57B but at the high end of its own history, and its share count fell 0.5% quarter over quarter to 1.57B — steady buybacks compounding per-share value. The dividend side of the haven argument is not a stretch. JPM's trailing twelve-month payout is now $6.15 per share, up 10.8% annually, with the most recent quarterly payment raised to $1.65 from $1.50 (ex-date October 6, 2026). Morgan Stanley's trailing payout is $4.15 per share, up 9.9% annually, after raising the quarterly dividend to $1.15 from $1.00 in July. Both companies have raised dividends in every year of the available record, which is exactly the income cushion the idea claims falling tech names cannot offer. Capital discipline reinforces the story: JPM's share count has fallen from roughly 3.97B in 2007 to 2.66B today, a relentless shrinkage that turns stable earnings into faster per-share growth. Full-year 2025 JPM net income was $57.0B on $182.4B of revenue, so the $21.2B Q2 2026 quarter represents a meaningful step up, not a one-off. With per the CNBC piece BNY popping more than 3% on strong earnings even on a weak tape, the pattern of banks beating expectations and being rewarded is at least established for the season's early reporters. On the trading plan itself: the entry rules were evaluated on real daily bars but never opened a position in the evaluated window — 1,236 daily bars for MS over 60 months, plus 24-month and 12-month windows, all with zero triggers. That makes this a watch-list setup rather than an active signal,…

MS Return on equityReturn on equity trend from CommonQuant fundamentals/XBRL data; -85.0% from first to latest point.
MeasureValue
2007-11-300.09754688877405236%
2008-11-300.03358186933170703%
2009-03-31-0.003597049200317028%
2009-06-300.0031983857811359634%
2009-09-300.016332606960236466%
2009-12-310.028829677861549004%
2009-12-310.013215387251542152%
2010-03-310.03679761312779711%
2010-06-300.038421578813243686%
2010-09-300.0023032562064842815%
2010-12-310.08220447116813201%
2010-12-310.014612574504902904%
Latest Value0.014612574504902904%
Change Pct-85.01994816180147%
TickerMS
Timeframereported periods
MS RevenueRevenue trend from CommonQuant fundamentals/XBRL data; -70.8% from first to latest point.
MeasureValue
2007-11-30$26478000000
2008-11-30$22140000000
2009-03-31$2934000000
2009-06-30$5196000000
2009-09-30$8468000000
2009-12-31$23280000000
2009-12-31$6836000000
2010-03-31$8986000000
2010-06-30$7925000000
2010-09-30$6733000000
2010-12-31$31230000000
2010-12-31$7743000000
Latest Value$7743000000
Change Pct$-70.75685474733741
TickerMS
Timeframereported periods

Scores

  • Conviction score breakdown: 50
  • Thesis support: 65
  • Trade readiness: 40
  • Risk quality: 45
  • Trigger proximity: 40
  • Fundamentals trend: 60

Watch items

  • JPM — Close vs. nearest resistance
  • JPM — Close vs. SMA (20)
  • JPM — Close vs. nearest support
  • JPM — Insider net open-market activity
  • MS — Close vs. SMA (20)
  • MS — Close vs. nearest support
  • MS — Insider net open-market activity
  • JPM — Next dividend event
  • BK — Price data availability
Unlock full analysis

Key details

BKJPMMS1d#canonical-demand#cluster-version:1#direction:bullish#entity-kind:instrument#entity:BK#entity:JPM#entity:MS#horizon:unspecified#intent:research#symbol:BK#symbol:JPM#symbol:MS

Community

0
Upvotes
4
Views
0
Copies
0
Cosigns

News sources

Related

CommonQuant logoCommonQuant.aiA Replikate Business
Loading…