What You Will Learn
Most portfolio construction tools give you a pie chart and call it a day. Journely runs a full investment team workflow — sector analysts assess valuations, a market strategist evaluates the regime, a security selector picks individual holdings with detailed thesis, and a risk manager validates every constraint before a single trade is placed.
This tutorial walks through a real portfolio built on February 15, 2026. A user asked to deploy $100,000 into AI and semiconductors with moderate risk tolerance. What happened next involved six specialized agents, one portfolio rejection, a forced rebalance, and a final approved allocation of six stocks plus a 22% cash reserve. Every number below is actual Journely output.
- Intent parsing — Journely extracts capital, currency, sectors, risk tolerance, and constraints from natural language
- Market regime analysis — macro conditions, sector valuations, rotation signals, and entry timing assessed before picking any stocks
- Security selection — 6 holdings chosen with individual investment thesis, quality metrics (ROIC, ROE, FCF margins), and screened-out value traps
- Risk validation — the risk manager rejected the first proposal for insufficient cash reserves and excessive concentration, forcing a revision
- Execution — final approved portfolio with exact share counts, compliance checks passed, and trades placed
Building a $100K AI/Semiconductor Portfolio
State Your Goal
You do not need to know asset allocation theory or sector weights. Just tell Journely what you want to invest in, how much capital you have, and your risk tolerance. The system parses your intent and extracts the parameters automatically.
Total Capital: $100,000 USD. Primary Sectors: AI, Semiconductors. Risk Tolerance: Moderate. Position Status: Greenfield (no existing positions).
Moderate risk tolerance means: blend of growth-oriented and more stable names, diversification within sector to avoid concentration risk, focus on profitable companies, and position size limits to control downside exposure.
Market Context Before Stock Picking
Before selecting any holdings, Journely assesses current market conditions. The sector analyst evaluates valuations across the AI/semiconductor value chain. The market strategist determines the current regime — risk-on, neutral, or risk-off — which directly impacts how much capital to deploy and how quickly.
Economic Context: GDP growth 4.4%, Manufacturing PMI 52.4 (expanding), Fed funds 3.75% (accommodative for growth). Market breadth exceptional at 9.5x advance/decline ratio despite Fear & Greed at 36.3 — a contrarian buy signal.
Electronic Technology (Semiconductors): P/E 63.2x, +5.88% YTD, +63.57% 1-year. Expensive, momentum slowing. Technology Services (AI Software): P/E 52.5x, -10.15% YTD. Oversold — unjustifiably sold off despite strong fundamentals.
Recommended subsector allocation: AI Software 22-25% (oversold, mean reversion opportunity), Chip Designers 20-23% (core AI exposure), Manufacturers 18-22% (margin expansion play), Equipment 15-18% (picks-and-shovels), Complementary 12-15% (diversification buffer).
The economy is growing at 4.4% but investors are rotating OUT of growth/tech INTO defensive assets. VIX at 20.6 (85th percentile), QQQ oversold (RSI 36.63), tech-specific volatility at 26.87. Bonds rallying (TLT +2.51% weekly), dollar strengthening (DXY 118.24) — classic risk-off positioning despite strong fundamentals.
Recommendation: Deploy via SCALED ENTRY (30-40% now, 30-40% on weakness, 20-30% on stabilization) rather than lump sum. Maintain 20-25% defensive cash buffer. Favor profitable, cash-generative names over speculative growth.
Why regime assessment matters
Most robo-advisors skip this step entirely — they pick stocks without considering whether the market environment favors aggressive or defensive positioning. Journely's market strategist determined this was a NEUTRAL regime, which reduced the recommended equity allocation from 80-85% to 70-75% and added a mandatory 20% cash reserve. That decision protected the portfolio from immediate volatility risk.
Security Selection with Individual Thesis
The security selector screens candidates using ROIC, ROE, FCF margins, PEG ratios, and analyst consensus. Each selected stock gets a detailed investment thesis explaining why it was chosen and what catalysts to watch. Crucially, the system also shows which stocks were screened out and why.
MSFT ($14,849, 14.9%) — AI software leader. P/E 25.1x, PEG 0.88, ROE 34.4%, net margin 39%. Copilot adoption driving cloud margin expansion. Most diversified big tech name.
NVDA ($14,990, 15.0%) — Dominant AI chip designer. P/E 45.3x but PEG 0.78 (undervalued for growth). ROE 107.4%, ROIC 97.1%, net margin 53%. 80%+ data center GPU market share.
AVGO ($14,958, 15.0%) — AI networking and custom silicon. P/E 68.3x but PEG 0.24 (extremely cheap relative to 291% EPS growth). FCF margin 42.1%. Hyperscaler partnerships unlocking.
LRCX ($13,896, 13.9%) — Semiconductor equipment. Highest ROIC in universe at 45.8%, ROE 65.6%. Leading supplier for advanced node production. Pricing power in supply-constrained environment.
KLAC ($11,713, 11.7%) — Chip inspection and metrology. ROE 100.7%. Inspection is less cyclical than equipment — a cost of production, not discretionary. AI-powered defect detection tools.
AMAT ($11,712, 11.7%) — Diversified equipment supplier covering 40%+ of fab process steps. P/E 36.3x, PEG 1.21 — most conservative valuation in the group. Broadest customer base (Samsung, Intel, TSMC).
Cash reserve: $15,882 (15.9%).
Stocks that were screened out
The system explicitly rejected AMD (P/E 78x with ROE only 7% — value trap), QCOM (declining earnings -48% YoY), Intel (negative profitability, FCF -9.4%), META (negative EPS growth -1.8%), and Micron (low FCF margin 11% vs peers). Knowing what was rejected and why is as valuable as knowing what was selected.
Risk Manager Rejects the First Proposal
This is what makes Journely different from a stock picker. The risk manager independently validates every portfolio against hard constraints — and has the authority to reject proposals that do not meet the criteria. The first version of this portfolio was rejected on three grounds.
Cash Reserve: CRITICAL VIOLATION. Current 15.9% vs required 20% minimum. Shortfall of $4,100. With a NEUTRAL market regime and VIX at 85th percentile, a 15.9% cash buffer is inadequate.
Sector Concentration: EXCESSIVE RISK. 82.2% of portfolio in a single mega-sector. If semiconductor/AI cycle peaks or demand reverses, 82% of portfolio suffers simultaneous losses.
Max Drawdown: -24.6% exceeds the -20% threshold for moderate risk tolerance. Sharpe ratio of 0.41 indicates inadequate risk-adjusted returns for a NEUTRAL regime.
REQUIRED: Increase cash to 20% minimum (+$4,100). Reduce equity allocation proportionally. Do not approve any trade that keeps cash below 20% given current market conditions.
Why the rejection matters
The security selector optimized for returns. The risk manager optimized for survival. These are different objectives, and the tension between them produces a better portfolio. The risk manager forced an additional $4,100 into cash reserves and flagged that a -24.6% max drawdown might be uncomfortable for moderate risk tolerance — information most portfolio tools never surface.
Revised Portfolio — Approved and Executed
After the rejection, the portfolio was automatically revised to meet all constraints. Each position was reduced proportionally, freeing $4,100 for the cash reserve. The compliance agent then verified every constraint before approving execution.
MSFT: 35 shares at $401.32 = $14,046 (14.0%). NVDA: 79 shares at $182.78 = $14,440 (14.4%). AVGO: 44 shares at $325.17 = $14,307 (14.3%). LRCX: 57 shares at $235.53 = $13,425 (13.4%). KLAC: 7 shares at $1,464.13 = $10,249 (10.2%). AMAT: 32 shares at $354.91 = $11,357 (11.4%). Cash Reserve: $22,175 (22.2%).
Compliance: Max position 14.4% (under 15% limit). Cash reserve 22.2% (exceeds 20% minimum). Budget 100% utilized. All 6 stocks rated Strong Buy or Buy. Average EPS growth +43% YoY. Portfolio volatility 12.3% annual.
All 6 orders executed successfully. Portfolio is live.
What Makes This Different
- Multiple agents with different objectives — the security selector optimizes for returns, the risk manager optimizes for survival, and the compliance agent enforces hard constraints. The tension between them produces a portfolio no single agent would build alone
- The portfolio was rejected and revised — this is not a chatbot giving you a list of stocks. The risk manager independently flagged that cash reserves were insufficient and forced a rebalance before any trade was executed
- Market regime drives allocation — the NEUTRAL regime assessment reduced equity from 85% to 78% and mandated scaled entry rather than lump sum. Context-aware allocation, not generic rules of thumb
- Value traps screened out — AMD (78x P/E, 7% ROE), Intel (negative profitability), and META (negative EPS growth) were explicitly rejected with reasoning. Knowing what NOT to buy is as important as knowing what to buy
- Every number is from a real execution — actual fill prices, real share counts, live compliance checks. This portfolio was built and executed on February 15, 2026 using real market data