Enter two tickers. Get the objective winner by 8 weighted factors from SEC. No hype. Pure math.
Loading…
Comparative analysis of CI and HUM reveals key differences across fundamental metrics.
CI shows stronger revenue growth (15.0%) compared to HUM (11.8%), indicating faster business expansion.
CI has higher operating margin (3.3%) than HUM (2.1%), suggesting better cost efficiency.
In terms of return on invested capital (ROIC), CI outperforms with 10.1% versus 7.3% for HUM, indicating superior capital efficiency.
CI generates more free cash flow margin (3.1%) than HUM (0.3%), highlighting stronger cash generation ability.
Based on our weighted algorithm, CI emerges as the winner with a score of 82 against 18 for the opponent.
Disclaimer: This analysis is for informational purposes only and does not constitute investment advice. Always perform your own research before making investment decisions.
Relative only — not absolute. Winner = more weighted points vs opponent.
Wᵢ — factor weight (sum = 1.0) · Nᵢ — normalized rank vs opponent (0=lose, 1=win, 0.5=tie) · Missing data → neutral 0.5 · Result: 0–100
Short, checkable takeaways from our fundamentals series. Full papers live on Substack — no forecasts, only SEC data run through the same formulas you just used.
Strong top-line growth is easy to celebrate and harder to cash. Across a 72-company, five-sector panel we asked a narrower question: where does 3-year revenue growth stay healthy while operating profit already runs well ahead of free cash flow — or while accruals push the Sloan Ratio above its own moderate threshold? Two independent screens, same filings, no forecasts. About a quarter of high-growth names show a wide operating-to-FCF margin gap; about one in six show elevated Sloan. Three companies trip both. Most of those names still read OK on DUEL's built-in Growth-vs-FCF check — not because either test is wrong, but because they answer different questions. A reusable detector for the next growth cycle, built only from public SEC data.
Full deep-dive on Substack →Every AI chip is made on someone else's machine. Inside the listed semiconductor capital-equipment oligopoly — AMAT, LRCX, KLAC, TER, ONTO — we asked a narrower question than the broad infrastructure maps: who wins a systematic head-to-head, who still has modeled room after high margins, and does resilience track the “obvious” cycle leader? Relative strength, DCF upside, and balance-sheet armor point to three different names. The round-robin top spot itself is not fixed: under production weights LRCX edges AMAT; under a quality-tilted scheme the ranking flips. Same eight SEC factors, transparent formula, more than one reasonable reading of a tightly matched group.
Full deep-dive on Substack →The picks-and-shovels map of AI infrastructure usually stops at the rack. Chips, networking, and rack-level power get the headlines — but every rack still has to plug into generation, the grid, and the contractors who build the interconnect. We ran that chain through the same DUEL lens as the prior infrastructure piece: ten US-listed suppliers across generation, electrical equipment, and T&D construction, plus the four hyperscale buyers as a control group. Growth and margin split into three uncorrelated stories. Resilience concentrates in equipment and contracting. And for three of four generation names, DUEL's own trailing FCF DCF hits its model floor — not a verdict on those utilities, but a clear limit of CFO—CAPEX math on rate-base balance sheets. Same filings, same formulas, a different layer of the buildout.
Full deep-dive on Substack →We ran 11 AI-infrastructure suppliers and 4 hyperscale buyers through DUEL's Battle, DCF, and Resilience reports — all from the same SEC filings, no analyst estimates, no macro forecast. The basket spans compute (NVDA, AMD, AVGO, MU), networking (ANET, CIEN), power & cooling (VRT, ETN, MOD, GEV), and storage (WDC). Nine of eleven names still grow revenue in double digits, but resilience does not follow the growth story. The most-discussed name in the cycle currently shows among the least remaining modeled room in DUEL's own DCF math — not a forecast, just a snapshot of what the filings already show. The full piece maps who is actually capturing the buildout right now, and how much of that capture is already priced in.
Read the full deep-dive →A single dramatic case — Oracle's 84% ROIC sitting next to a Sloan Ratio flag — raised an obvious question: is that the normal way a company fails a consistency check, or the rare exception dressed up as the norm? We ran 60 companies through the Resilience Report and asked: what is the base rate of each of the three internal conflicts (Growth vs FCF, ROIC vs Sloan, Debt vs Cash), and which one actually drives most penalties? The answer changes how you read a flagged note. Wilson confidence intervals, not Wald approximations — because rare events need the right statistics.
Read the full analysis →DUEL scores every duel on eight metrics with fixed weights. But financial ratios rarely cooperate with the assumption of independence. We ran a principal component analysis on 60 companies across technology, healthcare, industrials, and retail — and asked: how many statistically independent signals are actually encoded in the eight metrics, and how do they group? The answer has direct implications for how much of the 24—27% combined weight on correlated factors is really voting once, not twice. A methodological pilot worth extending.
Read the full piece →The weights behind DUEL's score are a reasonable expert starting point — not something backtested. So how much does the winner change if a reasonable person picks different, equally defensible weights? We tested this at three levels: a clean fragility number (gap ÷ 2), a Monte Carlo simulation (10,000 draws per duel), and Shapley-Shubik power indices. The results: most duels are stable, but AMZN vs WMT is fragile by an order of magnitude — and ROIC dominates the flip mechanism for that pair by a wide margin.
Read the full analysis →A DCF model that outputs $176.47 — precise to the cent — rests on a terminal value that makes up 60—80% of enterprise value, built on one unverifiable assumption about perpetual growth. We closed that gap without Monte Carlo: a three-point PERT estimate, using only the model's own published tiers. The result is a defensible range for every company — and a quiet side benefit: running this series doubled as an audit of our own model, surfacing two concrete improvements to the live tool.
Read the full piece →A duel only ever compares two companies. Classical relative valuation compares a company to a peer group, not to one arbitrary rival picked for a headline. We built a 2-round tournament across 12 semiconductor companies, using the Bradley-Terry model to estimate sector-wide rankings from paired comparisons — and measured exactly how much stability that gives you (H1, H2, H3 all held up). The result: a sector ranking that separates relative value, absolute value, and resilience — three different lenses, three different answers.
Read the full piece →Every valuation framework eventually collapses into one of two camps: intrinsic (DCF) or relative (multiples, peer ranking). They rest on different assumptions about what you can know — and they can point in opposite directions on the same pair. We walked through the theory and tested it on real SEC numbers (NVDA vs AMD, GOOGL vs META, AMZN vs WMT) so you can see when each lens is useful — and when it is not. A methodology write-up with real data.
Read the full piece →5 duels/day free. Upgrade when ready.