← Analysis page  ·  Joseph Carlson hub  ·  Research hub

Actionable insights — LIVE Discord Ask Me Anything

Not which names he likes, but how he sorts AI winners from losers in software, spots a priced-in bear case, and manages risk as a portfolio grows.
2026-SEP-15 · Qualtrim Studio — Investor Exchange · Joseph Carlson (live member Q&A) · ▶ Watch on Qualtrim · full analysis · transcript
How to read this page: each insight is a reusable procedure. The boxed line shows how it played out in this AMA. Timestamps are positions in the login-gated Qualtrim Studio recording (the link opens the video).

1:37 1. Prefer clean stories; demand KPIs for messy ones

The repeatable method
  1. For each company exposed to a new technology, list which segments it helps and which it hurts.
  2. If only one direction applies, it's a clean story. If both, it's messy — decide the KPIs that would prove the net effect and how many good quarters you need.
  3. Default to lower sizing on messy stories.
Here: NFLX clean; DIS (streaming vs cable) and ADBE messy; consumer INTU dicier than enterprise NOW/CRM.
Watch for

5:22 2. Check revenue, not the chart, before accepting a disruption story

The repeatable method
  1. Find the company the narrative says "should be destroyed."
  2. Compare its stock move with its revenue trend over the same period.
  3. If revenue still grows, the fall may be multiple compression from a high starting valuation, not disruption.
Here: NICE (AI-native customer-service target): stock "destroyed," revenue still growing; radiologists supposedly doomed, now in shortage.
Watch for

39:38 3. Price the bear case as if it already happened

The repeatable method
  1. Value the business segment by segment (sum of parts).
  2. Zero out or haircut the threatened segment.
  3. If the remaining segments are worth about the market cap, the bear case is already priced in.
Here: GOOGL: Cloud, YouTube and Waymo alone exceeded the ~$1.5T cap during the ChatGPT scare. Now applied to UBER, where a member's sum-of-parts puts mobility at ~85% of market cap.
Watch for

35:42 4. Put the disruptor and incumbent on the same unit

The repeatable method
  1. Pick a single operating metric both report (trips, users, orders).
  2. Convert to a vivid ratio (how long the incumbent takes to match the challenger's weekly volume).
  3. Track whether the absolute gap is widening or narrowing.
Here: UBER matches Waymo's weekly paid trips in ~13 minutes and adds "multiple Waymos" of trips each quarter.
Watch for

58:41 5. Diversify by risk factor, not by count

The repeatable method
  1. Hold at least ~7 positions — about 90% of single-company risk is diversified away by then.
  2. Group holdings by what could hurt them (rates, regulation, consumer spending, AI disruption); avoid many names sharing one factor.
  3. Weigh balance-sheet strength and business breadth; cap any one position's weight and trim if it grows too large.
  4. As wealth grows, park enough in ETFs to cover living needs, then invest the rest aggressively.
Here: Twenty regional banks = one rate bet; AMZN is "almost… like an ETF." He trimmed TXRH at top-two weight; with $25M he'd put $5–7M in ETFs (VTI) first.
Watch for

42:06 6. Judge heavy capex against how the build environment will change

The repeatable method
  1. Ask whether the inputs (land, permits, power) get easier or harder to secure over the next few years.
  2. If harder, early spending buys scarce position even at the cost of near-term FCF.
  3. Separate that from spending in an environment where inputs will get cheaper.
Here: META's ~$180–200B capex: data-center permitting and land near cities only get harder amid public backlash and possible political change.
Watch for

Methods distilled from the login-gated Qualtrim Studio video for personal study. Not investment advice.