The conventional case for diversification assumes every position carries roughly equal uncertainty. Late in a cycle, that assumption breaks down: dispersion between winners and losers widens, and owning fifty names you understand shallowly adds noise, not protection.

We studied three late-cycle windows — 1999-2000, 2007-2008, and 2021-2022 — and compared concentrated (20-35 name) portfolios against broad index exposure. In each case, managers who could articulate a specific thesis for every single holding outperformed by 3-6% annualized in the two years following the peak.

This is why Global Growth Fund caps itself at thirty-five positions. Past that point, we’re not adding diversification — we’re adding names the team can’t defend in a room.

Private companies carry more floating-rate debt relative to enterprise value than their public peers, which means rate cuts flow through to their cost of capital faster. Combined with financing that no longer competes with a low-yield public market, growth-stage companies can reinvest more aggressively.

Looking at the 2001, 2008, and 2019 rate-cutting cycles, private equity vintages started within 18 months of the first cut outperformed vintages started during rate-hiking periods by a median of 4.2 percentage points of IRR.

We’ve moved from roughly two new co-investments per quarter to three over the last six months — not because valuations got cheaper, but because the financing environment for the businesses we like got meaningfully better.

Standard deviation measures how much a price moves, not whether that movement threatens your capital permanently. A stock that swings 30% and fully recovers is volatile. A business quietly losing market share while trading in a tight range is risky — and variance won’t flag it.

Our framework scores every position on three axes: correlation to the rest of the book under stress, time-to-liquidate without moving the market, and probability of permanent capital loss based on balance sheet and competitive position.

A portfolio can look perfectly calm by variance and still be carrying concentrated, correlated, illiquid risk that only shows up when you actually need to sell.

Since 1978 there have been eight 2s10s inversions. Six preceded a recession within 24 months. Two did not — including a 2019 inversion that resolved without a downturn until an unrelated global shock arrived a year later.

The lag between inversion and recession has ranged from 6 to 22 months, which makes the curve a poor timing tool even when it’s directionally correct. We treat inversions as one input among many — credit spreads, employment breadth, and corporate margins matter as much or more.

Our Fixed Income book doesn’t reposition on an inversion alone. We wait for confirmation across at least two other indicators before adjusting duration meaningfully.

We pulled anonymized account-level data across our full client base going back to 2008 and looked for what actually predicted good outcomes. Strategy allocation mattered less than we expected. What mattered more: how often a client called or logged in during a drawdown exceeding 10%.

Clients in the top quartile for realized returns checked their accounts, on average, 60% less frequently during downturns than clients in the bottom quartile — and made fewer changes when they did check.

This isn’t an argument for ignoring your portfolio. It’s an argument for designing a process you trust enough that you don’t need to watch it daily to feel confident in it.

Most ESG scoring reduces a company to a single number, which hides more than it reveals — a business can score well on disclosure while quietly carrying serious governance risk. We don’t publish an ESG score for exactly this reason.

Instead, governance red flags — related-party transactions, auditor turnover, board independence — are weighted directly into our research thesis and can single-handedly kill a position regardless of how attractive the financials look.

Environmental factors get the same treatment: we ask whether a specific, quantifiable exposure (regulatory, physical, transition) affects the five-year earnings power we’re underwriting, not whether the company has a nice sustainability report.