Trading Signals

Executive summary

As of 2026-06-24 the market is in a mega-cap-tech & enterprise-software correction inside an otherwise richly-valued global market, driven by a Middle-East war (from ~27 Feb 2026), re-accelerating inflation, and a hawkish pivot (Fed on hold, market pricing a hike; ECB hiking).

• The broad market held up; SAP and MSFT did not. Trailing 12 months: S&P 500 +22.7%, MSCI ACWI +15.6%, Microsoft −17.8%, SAP −44%. Both underperformed the index they belong to. [83][84][28][7]
• The de-rating made them cheaper, not broken. MSFT forward P/E ≈ 19.8× (from ~30× two quarters earlier); SAP forward P/E ≈ 17.5–19.5×. Both still sit above Europe’s ~15.9× and at/above the S&P’s elevated 20.8×. [24][6][79][81]
• The cleanest contrast is cash generation. SAP’s FCF nearly doubled to €8.2 bn (FCF yield ≈ 5.3%, capex-light, rising). Microsoft’s FCF fell to $71.6 bn and is being squeezed by AI capex — Q3-FY26 FCF dropped to $15.8 bn with ~$190 bn of CY2026 capex guided (FCF yield ≈ 2.6% and falling). [1][20][22][23][26]
• The core question — “what do I give up vs. a global index fund?” Vanguard’s 10-yr model projects 4–5%/yr for US equities (ex-US 4.9–6.9%), with a 2.25% real, risk-free 10-yr TIPS yield as the hurdle. Neither stock is an obvious free lunch over a cheap global tracker, but SAP’s valuation reset and MSFT’s franchise quality each give a credible relative case. [78][72]

Part A — The companies

A1 · SAP SE (ETR/NYSE: SAP) — FY2025

World-dominant ERP vendor, ~58% through a licence→cloud shift (RISE/GROW with SAP, “Business AI”/Joule). AI strategy is monetisation-led and capex-light — it rents compute rather than building it.

Q1 2026: cloud revenue €5.962 bn (+27% cc), current cloud backlog €21.9 bn (+25% cc), non-IFRS EPS €1.72 — but total revenue missed and the stock sold off; management flagged decelerating cloud growth into 2026. Valuation: ~€134 (Xetra), market cap ~€156 bn/$178 bn; forward P/E ~17.5× NTM; EV/EBITDA ~12.5–15.5× (~30% below 10-yr median); FCF yield ~5.3%; dividend €2.50 (~1.9%); net cash ~$2.5 bn. Consensus: Buy. [1][2][4][5][6][12]

A2 · Microsoft Corp. (NASDAQ: MSFT) — FY2025 (ended 30 Jun 2025)

The most direct “AI infrastructure + application” mega-cap (Azure + Copilot). Two new overhangs: the AI-capex cycle is turning a capital-light compounder capital-heavy, and an escalating FTC antitrust probe.

Q3 FY2026: revenue $82.9 bn (+18%), Azure +40% — but FCF compressed to $15.8 bn and management guided ~$190 bn CY2026 capex. Valuation: $372.99, market cap ~$2.73 tn; forward P/E ~19.8× (was ~30× two quarters earlier); EV/EBITDA ~15.3×; EV/Sales ~10.3×; PEG ~1.25; FCF yield ~2.6% and falling; dividend ~1.0% + ~0.7% buyback; net cash ~$54 bn; FY2025 capital return ~$42.5 bn (largest in the S&P 500). [19][20][21][22][24][25][30][31]

A3 · Total shareholder return (price + dividends)

*Both melted up then round-tripped: SAP returned +52% (2023) and +61% (2024) before collapsing in 2025–26. Reported “3-yr” snapshots taken near the peak are stale and excluded; 5-yr CAGRs are period-/date-sensitive (see “What I could not verify”). [7][14][28]

Part B — Relative to the stock market

Total world equity market cap ≈ $127–154 tn (US ~half). On that base Microsoft (~$2.73 tn) is ~1.8–2.1% of all listed equity on Earth; SAP (~$178 bn) is ~0.12% — one is a macro asset, the other a large-cap. [86][87]

†DAX single-stock weights are capped; the “16.2%” and “SAP+Siemens 18.98%” figures don’t fully reconcile — treat SAP as the dominant DAX constituent (exact weight flagged). Asymmetry: own ACWI/VT and you already hold ~3% MSFT; SAP barely moves a global tracker but dominates German/European indices — a concentration risk for DAX-heavy investors. [38][40][44][46][47][41][43]

B2 · Sector & peers — the 2026 software slump

Enterprise software de-rated hard in 2026: ServiceNow’s EV/EBITDA collapsed 47×→30×, Salesforce to ~14× forward P/E.

SAP and Microsoft sit in the attractive middle of the growth-vs-margin frontier (double-digit growth at GARP-ish multiples) and trade at a discount versus their own histories — though not versus the broad market. [48][49][50][51]

B3 · Style & factor exposure

• Beta: SAP 5-yr ~0.7–0.8, MSFT ~1.1 — but SAP’s 2024–26 round-trip means historical beta understates recent risk. [33][34]
• Drawdowns: now SAP −50% / MSFT −33%; 2022 rate shock SAP −24% / MSFT ~−28%. These are not defensives — they led the current losses.
• Yield: SAP ~1.9% > MSFT ~1.0% (+~0.7% buyback). Both quality-growth; SAP screens more GARP/quality.

B4 · Alternatives — “what do I give up vs. a global index fund?”

Answer: versus a global tracker (~4.5–5.5%/yr modelled), holding MSFT bets its AI franchise out-earns the index despite a 2.6%-and-falling FCF yield at ~20×; holding SAP bets a 5.3% FCF yield + the ECC migration catalyst + a 17–19× multiple beat the index. Either way the cost is concentration & single-name risk (antitrust, capex, FX) for a return edge that is plausible but not guaranteed — and you forgo the cheap pockets the model favours (EM, value, IG credit). [78][72][55][68][69][67][85]

Part C — Relative to the world economy

The regime is a stagflationary war shock — slower growth + higher inflation + a pivot back toward tightening — the worst mix for long-duration growth equities, which is why tech/software de-rated.

• Geography: SAP EMEA ~46% / Americas ~40% / APJ ~14% (more Europe-levered); MSFT US 51% / Intl 49% (balanced). FX: SAP’s constant-currency growth ran ~7–8 pts above reported in 2025–26.
• IT-capex cycle (tailwind): Gartner sees worldwide IT +13.5% to $6.31 tn and software +15.1% to $1.44 tn in 2026 — “the slowdown never came.”
• AI-capex bubble vs. productivity (the fault line): big-5 hyperscaler capex >$600 bn in 2026 (+36%), ~94% of operating cash flow after dividends/buybacks; critics flag understated depreciation, circular vendor-financing & ~$662 bn off-balance-sheet leases. Microsoft is squarely inside this; SAP is largely outside it — the single biggest difference between the two as macro bets.
• Idiosyncratic risks: SAP’s 2027/2030 ECC maintenance cliff (a migration tailwind); MSFT’s FTC antitrust probe (Azure/Copilot bundling, OpenAI; 70+ subpoenas — an overhang); EU AI Act fines from 2 Aug 2026 (up to 7% of turnover).

Goldman: GenAI could lift global GDP ~7% / +1.5pp productivity over a decade — but skeptics (Acemoglu) see ~1%, and Goldman’s own read is GenAI added “basically zero” to measured 2025 growth. Much of the infrastructure upside looks priced in; the application-layer monetisation (SAP Joule, M365 Copilot) less so. [52–77][93]

Part D — Synthesis

• SAP: above-average odds of beating ACWI over 3–5 yr if cloud growth re-accelerates off the ECC catalyst and the multiple re-rates from a 50%-off base; ~5.3% FCF yield + ~1.9% dividend + ~8–12% EPS growth ≈ low-double-digit base-case IRR, lower capex risk. Risks: European macro, decelerating backlog, FX.
• Microsoft: franchise quality & Azure/Copilot argue for index-beating 5-yr returns, but the 3-yr path is hostage to the AI-capex payoff and the FTC. A ~2.6%-and-falling FCF yield at ~20× leaves little margin for error.
• Both underperform a simple index fund if: rates stay higher-for-longer and multiples compress market-wide; the AI-capex cycle disappoints (hits MSFT hardest); IT budgets crack in a recession; or the cheap pockets (EM, value, ex-US) lead.
• Diversification: ACWI/VT already gives you ~3% MSFT + ~0.1% SAP plus everything else. SAP+MSFT together is a factor bet (quality-growth software), not a diversified portfolio — they fell together.
• Currency (EUR investor): MSFT adds USD risk you already carry via any global tracker; SAP is EUR-native. EUR→USD hedging currently earns a small positive carry (USD rates > EUR rates).

Table 1 — SAP vs. MSFT vs. the market

Table 2 — Macro scenario matrix (illustrative 3-yr annualised total return — NOT a forecast)

Reading it: SAP’s distribution is narrower (capex-light + depressed starting valuation); Microsoft’s is wider and skewed to the capex outcome. In the bear case a diversified global index meaningfully outperforms both single names — the core argument against over-concentrating.

What I could not verify

• Exact current SAP market cap (sources span $178–200 bn by date/ADR-vs-ordinary).
• Precise SAP & MSFT 5-yr total-return CAGRs (period-/as-of-date sensitive; many snapshots were pre-correction).
• SAP’s exact weight in MSCI World/ACWI (not in ACWI top-10; ~0.1–0.3% is an estimate) and exact DAX weight.
• Microsoft’s forward FCF / FCF yield (depends entirely on CY2026 capex pacing; ~$190 bn guide).
• Precise S&P 500 YTD-2026 return (have trailing-12-mo +22.7% and ACWI YTD −4.9%).
• Table 2 returns & probabilities are illustrative, not modelled forecasts.
• Method limit: several figures came from secondary aggregators (stockanalysis, financecharts, tradingeconomics); confirm headline numbers against SAP’s 2025 Integrated Report/20-F and Microsoft’s 10-K/10-Q before acting.