The WBD call we got wrong

We graded Warner Bros. Discovery a Deep Dive in April. It was down 8.7% twenty-nine days later while the market was up. SPY returned +1.1% over the same window, so the call cost roughly ten points of net alpha. This is the post-mortem, because a methodology you can only see when it’s winning isn’t a methodology.

The thesis we wrote

The bull case was a deleveraging story: strong free-cash-flow conversion paying down a large debt load, a streaming segment crossing into profitability, and a sum-of-the-parts that screened well below the value of the pieces. The grade-cards liked it — Balance sheet improving, Execution credible on cost discipline, Valuation genuinely cheap, with a dated catalyst around a strategic review.

On paper it was a reasonable Deep Dive. In practice we made two mistakes.

What the model missed

The quantitative side over-weighted the trajectory of deleveraging and under-weighted its sensitivity to a single revenue line. The linear-TV decline wasn’t new information, but the model treated it as a slow, predictable bleak. It was not slow. An acceleration in cord-cutting compressed the cash flow that the entire deleveraging thesis depended on, and the sum-of-the-parts re-rated with it. The error wasn’t the inputs; it was assuming the worst-declining segment would decline politely.

What the analyst missed

Ender’s memo cited the catalyst — the strategic review — as a tie-breaker. The analyst read it as a near-term value-unlock. It was better read as a source of uncertainty: a review that could just as easily delay decisions as accelerate them. We let a catalyst that cut both ways get scored as if it only cut one.

The catalyst was real. We mis-signed it. “Something will happen” is not the same as “something good will happen.”

What we changed

Two rubric changes shipped after this call:

1. Segment-sensitivity stress. For any deleveraging or sum-of-the-parts thesis, the Balance-sheet and Valuation grades now carry an explicit stress against the single worst-trending revenue segment. If the thesis breaks when that line declines 20% faster than consensus, the grade is capped.
2. Catalyst sign discipline. A catalyst only improves the Catalyst grade if its direction is constrained. An open-ended “strategic review” now scores as elevated uncertainty, not a tailwind, unless the structure of the event limits the downside.

What we didn’t change

We did not change the verdict in hindsight or quietly delete it from Receipts. It sits in the table with its negative net alpha next to the calls that worked. We also didn’t conclude the process was broken — the grade-cards did their job, two specific weights were wrong, and we fixed the weights. The point of a post-mortem is to be precise about which part failed, so you change that part and not your nerve.

---

Research, not advice. This is a retrospective on a published call and not a current recommendation on any security.