Middle Coast Investing Q1 2026 Portfolio Activity

Booking shares fell as much as 32% from last year's highs and 25% year to date. The article says GitLab has performed poorly since its 2021 IPO and that the author may have bought Capital One too early at $210–$220; Capital One took over Discover a year earlier.

Analysis

The price weakness is a signal to test each thesis, not evidence that all three are cheap. For BKNG, the key distinction is cyclical demand softness versus a durable shift in customer acquisition economics: if paid search or metasearch costs rise faster than bookings and take-rate, earnings power can deteriorate even as travel volumes hold up. Conversely, a selloff driven mainly by positioning would leave the platform’s cash generation intact. Watch room-night growth, gross bookings, marketing expense, and forward guidance over the next 1–3 months; over 6–18 months, substitution by direct channels or alternative lodging is the structural risk.

GTLB’s weak trading history should not be used as a valuation thesis. The relevant question is whether AI features expand paid usage and retention or make software development tools easier to substitute. Verify net retention, large-customer growth, and operating leverage before treating a lower share price as an entry signal.

For COF, the second-order exposure is execution across the combined card and payments businesses, alongside credit losses and funding costs. A favorable integration narrative is not enough if charge-offs or delinquencies worsen. Near-term market moves may reflect rates and consumer-credit concerns; longer-term value depends on realized synergies without sacrificing underwriting discipline. There is insufficient information here to establish mispricing or quantify upside.

Key Decisions for Investors

• BKNG: Do not average down solely on drawdown. Reassess after the next results; turn constructive only if booking growth and marketing efficiency remain resilient. Falsifier: guidance cuts alongside rising acquisition costs.
• GTLB: Keep on a watchlist rather than initiate on IPO-era price history. Require evidence of stable retention and monetizable AI adoption; worsening retention or slower large-customer growth would invalidate the recovery case.
• COF: Treat as an integration-and-credit monitor, not a confirmed value entry. Track delinquency and charge-off trends, funding costs, and reported integration benefits; deteriorating credit metrics or absent synergy evidence argues against adding.
• No high-conviction trade is justified from this limited information. Revisit sizing only after reviewing current valuation, earnings expectations, and the next company disclosures.