The Compound and Friends

Among the top 20 US e-commerce sites today, 15 are legacy brick-and-mortar retailers that built their own websites, and only 5 are pure-play internet companies — directly contradicting the 1999-2000 consensus that online-native players would dominate retail.

Software stocks broadly were re-rated lower due to AI disruption fears, but Matt Ankrim argues the market is conflating high-consequence vertical SaaS (government, regulated industries) with low-consequence horizontal SaaS, creating mispriced opportunities in the former.

Toast has surpassed Block/Square and Fiserv/Clover to become the dominant restaurant POS platform with 150,000 locations and is expanding into hotel chains, with switching costs so high that Josh Brown argues the AI disruption thesis is wrong for this category.

AWS revenue growth re-accelerated to 28% year-over-year in the reported quarter — the fastest in 15 quarters — at a $150 billion annualized run rate, which Amazon's CEO Andy Jassy described as 'very unusual for a business to grow this fast on a base this large.'

An analyst team at Janus sold Fastenal after identifying an earnings miss, but the stock still rose 55% while the market fell 15% — and then delivered a 19-fold return from the sale price, demonstrating the severe cost of exiting high-quality businesses on short-term thesis breaks.

Fastenal's competitive moat derives from the fact that its products — nuts, bolts, and fasteners — represent less than 3% of total project cost for customers, making the service value (avoiding labor stoppages) far outweigh the price of the product itself.

The average stock holding period has collapsed from eight years in the 1960s to five and a half months today, structurally preventing most investors from capturing compounding returns that only materialize over decades.

A JP Morgan research report found that 40% of all stocks in the Russell 3000 experience a catastrophic 70%+ decline from which they never recover, meaning avoiding permanent losers is as important as finding winners.

Warren Buffett generated 98-99% of his total wealth after age 65, illustrating that the overwhelming majority of compounding returns accumulate in the final years of a long holding period.

High-quality stocks deliver higher returns with lower risk — a documented anomaly that GMO's Jeremy Grantham calls 'the weirdest inefficiency in the market' — contradicting the foundational finance principle that higher risk earns higher return.

84% of the top 50 companies responsible for 50% of all net stock market wealth creation were high-quality businesses by fundamental measures, validating a quality-first screening approach.

Only 4% of all stocks from 1926 to 2022 were responsible for 100% of net wealth creation in the market, making stock selection an extreme long-tail problem.

Nvidia will reach a $10 trillion market cap by end of the decade, growing revenue at roughly 15% per year for five years (at minimum doubling), with CUDA platform lock-in making large-scale AI workload migration practically impossible for the foreseeable future.

Broadcom is on track for 60% revenue growth in the current fiscal year with industry-leading profitability, and remains a core holding as the dominant provider of custom silicon and data center networking for hyperscalers.

Healthcare stocks are priced as if there is nearly zero probability the sector outperforms over the next five years; Parker estimates the true probability at 30-40%, creating a large mispricing given AI adoption potential and steady demand dynamics.

On average, spin-off companies outperform their industry peers by a statistically significant margin over two years while the remaining parent slightly underperforms — a pattern robust across sectors, arguing for owning the spin-co over the remain-co.

Sechan and Parker both prefer buying GP equity (e.g., Blackstone, KKR) over private credit LP positions, arguing that at current pricing LP investors get no discount to NAV while equity holders benefit from the upside of a fundraising recovery.

Blackstone's gross private credit fundraising dropped roughly 42% sequentially, from $3.3 billion to $1.9 billion in a single quarter — reflecting investor pause, though spreads remain contained and net flows are not negative.

The semiconductor ETF SMH recently posted its best-ever rolling 16-day return — a historically unprecedented pace of appreciation that Parker flagged as extreme even while acknowledging the fundamental rationale behind it.

ServiceNow grew its $5M+ ACV customer count from 434 to 630 with 97% retention and 22% revenue growth, yet the stock dropped 18% on the print — evidence that the software de-rating is driven by uncertainty about 2030 margin outcomes, not near-term fundamentals.

Street consensus expects S&P 500 earnings to grow 18% this year and 17% next, with the tech sector projected at 43% growth this year and 25% next — implying AI productivity gains are already embedded in forward estimates, not merely speculated.

45% of the entire S&P 500's Q1 2026 year-over-year earnings growth was attributable to just Micron and Nvidia, making index-level EPS highly concentrated and vulnerable if either stock's trajectory decelerates.

In software sector rebounds, the historically higher-returning strategy is to own expensive, fast-growing names like Cadence and Synopsys rather than cheap, slow-growing ones, because the premium price reflects lower technological disruption risk.

Software companies will first miss on gross margins and then subsequently miss on revenues, because analysts have modeled flat gross margins through 2027-2029 without accounting for new AI-tool spending that enterprise CTOs will eventually demand discounts on or build around.