Van Tuyl Companies · Private Equity
Every position graded on each metric, year over year, counted across the portfolio. This is the roll-up the workbook draws as pie charts on its Output Table.
Year-over-year change for each position, graded on the workbook's red / yellow / green thresholds. Price columns compare the holding against the S&P 500 and its sector ETF over the trailing year.
| Company | Inception | Sector | Revenue | EBITDA | Margin | Multiple | Equity value | Price | vs S&P 500 1Y | vs sector 1Y |
|---|
Each position at four points in time: the last twelve months as of inception, as of a year ago, as of today, and the next twelve months on consensus estimates.
Grading. Growth metrics (revenue, EBITDA, TEV, equity value, price) grade green at ≥ 5%, yellow at ≥ 0%, red below 0%. Valuation metrics (multiple, P/E) invert: green below 0%, yellow at ≥ 0%, red at ≥ 15%. Share count and net leverage grade green when falling, red when rising. Relative price performance grades green at ≥ +2 points, yellow within ±2 points, red below −2 points.
Ratio rows are graded in basis points. EBITDA margin, net income margin and FCF conversion are graded on the absolute move — green at ≥ +50 bps, yellow within ±50 bps, red below −50 bps — and the separate delta column is left blank for them, since it would repeat the same figure. A relative percentage on a margin ("up 3.8%") divides one ratio by another and carries no meaning; the workbook applied one to margin while already grading FCF conversion on the absolute change, and this makes the three consistent.
Units. Dollars in millions, share counts in millions, unless noted. NM marks values the workbook treats as not meaningful (typically net leverage where the company holds net cash); NA marks values it does not compute.
Currency. Everything is shown in USD. Adyen, ASML and Ferrovial report in EUR and TSMC in TWD; each figure is converted at the rate for the date it is struck on — the inception column at that position's entry date, the prior-year column a year back, the current column today. Because the rate moves between those dates, growth rates here are USD growth rates and will differ from the local-currency figures in the workbook. Ratios — margins, multiples, P/E, FCF conversion, dividend yield — are unaffected, since numerator and denominator share a currency. Rates are ECB reference rates (Taiwan dollar excepted, which the ECB does not publish).
Positions held through a US line. ASML, Ferrovial and TSMC are held via their US listings, so they quote in USD while still reporting in EUR or TWD — statements convert, the price does not. For TSMC the ADR also represents five ordinary shares while the share count is ordinary shares, so price × share count does not tie to equity value for that one position, and its dividend yield is suppressed because the dividend is per ordinary share in TWD.
Roster. Matches the High Conviction Model on page 3 of the Investment Committee notes dated 8/11/2026. The SOXX position is held but not shown: it is an ETF with no revenue, EBITDA or margins, so the tearsheet does not apply to it.
Inception. The "at inception" column is fixed to each position's entry date and does not change between refreshes.
P/E and the multiple are computed here, not taken from Capital IQ. Its API exposes no P/E or EV/EBITDA field, so both are built from primitives. P/E is price ÷ diluted EPS in every column, trailing and forward alike, so the NTM-vs-LTM cell measures a change in valuation rather than a change in construction. Capital IQ also publishes a normalized diluted EPS series; it is not used, because its forward consensus — despite carrying the same word — tracks reported EPS, and pairing the two would overstate expected growth by 25 points or more on most positions. The multiple keeps one unavoidable difference: trailing is against reported EBITDA while NTM is against an estimate, and there is no forward equivalent of a reported figure.
CAGR. Compounds from the inception value to the current LTM value, over the actual elapsed time. This differs from the workbook, which ends its RATE() at the prior-year column and so leaves out the most recent year. Only level figures carry one — compounding a margin or a multiple is not meaningful — and it is left blank where either endpoint is zero or negative, since a rate through a sign change would be fiction rather than a number. Read it with care on positions held less than a year: annualising a few months magnifies whatever happened in them.