Big banks return cash as Big Tech gets punished for spending — rotate into JPMorgan and Goldman

Verdict: The rotation thesis is real, but the evidence isn't — wait for the trigger

The strongest point for this idea is the payout evidence: Goldman's trailing 12-month dividend stands at $18.00 (up 38.5% from $13.00), JPMorgan's at $6.00 with return on equity improving 1.1 points to 5.6% in Q2, and both were stress-test validated per the June 24, 2026 CNBC report — regulators effectively signed off on the capital plans. The strongest point against is that the entire quantitative case rests on one completed trade over a 9-month window (6.0% return, 2.4% drawdown), a statistically meaningless sample, while the short leg itself is questionable: Microsoft sports a 67.9% gross margin, 30.2% return on equity, and $67.0B in free cash flow against $116.0B in capex, and the strategy's own exits sit dangerously close — XLF needs only a 0.47-point move above 57.72 plus a 0.59-point moving-average gap to fire the trend exit almost immediately after entry. Net insider selling across all three names for the June 30, 2026 filing period (−$29.3M GS, −$6.6M JPM, −$27.0M MSFT) further undercuts the shareholder-floor narrative, though those filings are point-in-time, not current. The verdict would flip if JPM's October quarterly report confirmed continued return-on-equity improvement and buyback-driven share reduction alongside a fresh ownership cycle showing insider selling turning to net buying. Until the XLK RSI (50.5) crosses below 50 to arm the short leg, this is an alert-on, not an act-on, idea.

Trade now: the financials leg is armed, tech is one RSI point away

**Action today: set alerts, not orders — the financials leg is live, the short leg is not.** On the XLF long setup, every entry condition is currently satisfied: XLF's 14-day RSI sits at 44.2 (needs below 50), the last close is 57.25 (needs above zero), and the 20-day Donchian upper band at 57.65 clears its threshold. The tech side is the sticking point — XLK's RSI is 50.5, just 0.5 points above the below-50 trigger, and MSFT's RSI is 50.7, 0.7 points above the same bar. GS and JPM sit in the same near-miss zone (RSI 49.5 and 50.5). Until the short leg confirms, 'wait' means exactly that: no position yet, alerts set on XLK and MSFT at RSI 50. **Once triggered, the risk math is fixed by the strategy's own exits.** The rules cut any position that falls 2.3% from entry and take profit at a 4.7% gain — roughly 2:1 reward-to-risk per leg. Position sizing is fixed-risk at 2.3% of the book per trade with a 20% cap on any single position. The signal exit is an XLF close crossing above its 20-day average (currently 57.72, with XLF closing 57.25 — only 0.47 away) combined with the 20-day average crossing the 50-day; both are flagged as near right now, so if you enter, understand the trend exit can come quickly. **The evidence base supports acting when it triggers, not chasing.** The completed backtest on the 9-month window traded once and returned 6.0% with a maximum drawdown of 2.4% — a single-trade sample, so treat it as a favorable read on the mechanics rather than a promise of frequency. The idea's thesis — banks passing stress reviews and hiking payouts while tech gets punished for spending — is consistent with the live tape: JPM's return on equity improved by 1.1 percentage points to 5.6% in the latest quarter, while MSFT trades 8.6% below its range high. Note one parameter caveat: the robustness evaluation ran out of its time budget, so no alternative parameter setup was recommended — trade the rules exactly as published. **Concretely, your checklist for today:** (1) alert on XLK and MSFT daily RSI crossing below 50; (2) alert on XLF at 57.72 — a close above it, while the entry is active, is also your trend-exit warning; (3) if entries confirm, size to risk 2.3% of the portfolio per leg, stops at −2.3%, targets at +4.7%, and re-check the 20/50-day average gap on XLF (currently 0.59 points) before assuming weeks of runway.

Stress tests, buybacks, and a rotation window that is actually open

The bull case rests on a rare alignment of policy tailwind, shareholder-return momentum, and visible sentiment rotation. Start with the policy leg: per the CNBC report of June 24, 2026, JPMorgan and Goldman Sachs passed their Fed stress tests and immediately followed with a $50B JPM buyback and a Goldman dividend increase — the idea argues this puts a hard floor under bank valuations, and the payout data backs it up. JPMorgan's trailing 12-month dividend is now $6.00 per share (up from $5.30 a year ago, a 13.2% annualized hike across 2025-2026), and Goldman's trailing 12-month payout is $18.00, nearly 38% above its prior 12-month rate of $13.00. These are not token gestures; they are capital plans validated by the stress test, meaning regulators effectively signed off on the balance-sheet strength the thesis is betting on. The rotation leg has teeth in the numbers. Per the MarketWatch piece of June 25, 2026, Microsoft is suffering a historic June rout as investors balk at heavy capex — and the idea argues the market is losing patience with Big Tech spending. On the fundamentals, Microsoft's FY2026 shows $331.8B in revenue with a 46.8% operating margin and $67.0B in free cash flow, yet it is spending $116.0B on capex alone, a figure that dwarfs free cash flow after tax and is exactly the kind of spend the market is punishing. Meanwhile, Goldman's most recent quarterly net income of $6.63B and JPMorgan's $21.16B in Q2 2026 show the banks are not just returning cash but actually growing earnings into the rotation. The completed backtest on the XLF/XLK long-short pair is the strongest quantitative support: over 182 daily bars in a 9-month window ending late 2025, the strategy returned 6.0% on a single trade with a 100% win rate and a maximum drawdown of just 2.4%. That is a tight risk profile for a sector-rotation trade — the trade did what the thesis predicts, and the drawdown stayed contained even while the pair was live. Note the caveats in scope: exits were filled on daily trigger bars rather than intrabar data, so stop-loss and take-profit fills are approximate and the drawdown figure is coarse, and the 60-, 24-, and 12-month windows could not be evaluated due to market-data gaps in XLF and GS daily coverage. Even granting those limitations, the one completed test matched the directional thesis. The cross-name ownership data adds a mild sentiment tailwind rather than a headwind for the thesis itself: institutional filing coverage as of the June 30, 2026 period shows net open-market insider selling across GS (-$29.3M), JPM (-$6.6M), and MSFT (-$27.0M), which is consistent with a broad de-risking posture rather than concentrated conviction in any single leg — the kind of environment where a defensive, cash-returning financial complex outperforms a high-beta, spending-heavy tech complex, exactly the rotation the idea is positioned for. With Goldman's ROE at 13.7% and JPMorgan's at 15.7% — both in the top 15-20% of their Financials peer group — and Microsoft's elevated capex cycle still pressuring sentiment, the setup the idea describes is visible in the data: cash-rich…

Scores

• Conviction score breakdown: 51
• Thesis support: 68
• Trade readiness: 45
• Risk quality: 55
• Backtest evidence: 30
• Fundamentals trend: 58

Watch items

• XLK — RSI (14)
• MSFT — RSI (14)
• XLF — Close vs 20-day average
• GS — Ex-dividend date
• JPM — Return on equity
• GS — Insider net open-market value
• MSFT — Insider net open-market value
• JPM — Shares outstanding