Research source kit

13 reviewed relationships preserve legal security classes, ownership form, exclusions, evidence scope, and observation dates.

812,004,746 NVIDIA shares were reported in the June 17 relationship; 91.4% were held indirectly.

19.3% had reviewed SEC position evidence; source reach was the first blocker for 77.5%.

9 of 9 matched founders had voting power above disclosed share-count ownership; Alphabet remains explicitly unmatched.

2 of 10 reviewed companies have a hard time-based companywide sunset; other mechanisms depend on events, thresholds, transfers, or holder votes.

The highest reviewed observation was 60.8% of total company voting power; 1 observation exceeded 50%.

4 of 10 reviewed founders exceeded 50% of total voting power; the highest observation was 82.4%.

5 increased, 4 decreased, and 1 was unchanged; the study does not assign transaction causes.

1,503 of 20,834 candidates (7.2%) were simple direct non-zero positions; footnote review accounted for 50.6% of candidates.

45,879 of 48,912 sale candidates (93.8%) carried a footnote-review blocker; 6,808 of 7,573 footnote-only cases used weighted-average or price-range language.

6 source patterns appear across 13 relationships; 4 relationships use multiple source families.

At $158.96, the reverse DCF requires about 37.19% annual revenue growth and $769.1B of 2035 revenue; 86.1% of modeled enterprise value comes from terminal value.

At approximately $300M of 2027 FCF applied to net debt, the 2.9x target requires about $703M of adjusted EBITDA, 8.22% above the $650M transaction base.

At the 76.75M middle hurdle, 65.5M of ShotSpotter revenue requires about 563 SafePointe lane-years and roughly 760 year-end live lanes under the generous 365-lane starting assumption.

Grizzly Bulls uses 50–60% as a normalized economic underwriting range, not a claimed audited GAAP margin.

59 of 61 companies (96.7%) normalized under the reviewed 13-metric annual-financials schema; the two remaining cases stopped at a CAD currency-unit boundary and an unresolved issuer-identity boundary.

3 of 9 reviewed companies had EV withheld despite complete net-debt review: 2 for noncontrolling interests and 1 for preferred equity; 6 were compatible.

The median disclosure lag was 6 calendar days; 5 of 21 observations (23.8%) were at least 10 days, with a 17-day maximum.

Four of five reviewed issuers stayed within a one-day lag band; 13 of 16 observations (81.3%) were disclosed in 5–7 days, while Netflix ranged from 17 to 23 days.

Cohort cash capex rose to 81.81% of operating cash flow from 60.80%, a +21.01 point change.

The three disclosed campuses imply a 12.56% aggregate midpoint site-NOI yield on development cost; Macquarie preferred capital compounds at 12.75%.

$227.475B of repurchases were observed across seven covered issuers; 5 of 6 positive-buyback issuers also reduced diluted shares.

Overall top-10 issuer market weight is 40.45%. The top-10 net-income cohort gap is -0.53 percentage points, while the top-10 capex cohort gap is -30.79 points.

June 2025 top-10 weight was 38.15%, the March 2026 cap-minus-earnings gap was 4.7 points, the June 2026 forward-P/E spread was 2.0x, and the forward-return correlation fell to -0.04 when 1965 was excluded.

The top 10 account for 26.32% of covered revenue and 50.68% of covered capital spending; Financials account for 74.79% of covered liabilities.

Starting concentration and subsequent decade return had a -0.72 sample correlation; excluding June 1965 reduced it to -0.04.

Top-10 concentration fell from 38.24% in 1965 to 17.81% in 2015, then rose 20.34 percentage points to 38.15% in 2025.

Top-10 trailing-earnings share rose from 25.5% to 33.2%, an increase of 7.7 percentage points, while the cap-minus-earnings gap narrowed to 4.7 points.

The top-10 minus remaining-company forward-P/E spread narrowed from 10.1x to 2.0x, while remaining-company P/E rose 1.3x.

The top 10 security lines were 38.09% of index weight; the conservative AI-linked cohort was 29.46%, or 77.3% of top-10 weight.

The May 2023 spread averaged -1.73 percentage points and reconstructs to 71.4% recession probability for May 2024; NBER did not classify that target month as recession as of the study cutoff.

With 100 independent zero-edge trials, the median winning Sharpe falls from 2.48 after one year to 0.78 after ten years; 1,000 ten-year trials push the median winner back to 1.01.

A true Sharpe of 1 has a 90.8% chance of at least one three-month losing streak over five years and a 56.8% chance of at least one four-month streak.

A true Sharpe of 1 implies a 38.64% modeled chance of negative one-month excess return and 15.87% over one year under the stated assumptions.