Actionable insights — Portfolio Update, Double Edition: QNST + J

1. Treat a price above the analyst target as a stale-model problem, not a sell signal

1. When a holding trades through the consensus price target, first ask when the target was set and what it was set on. A target published before a step-change in earnings is an artefact of the old model, not a judgment on the new one.
2. Measure the size of the surprise. A modest beat lets analysts tweak; a beat large enough to break the model — Haymaker's threshold here is a 42.7% revenue beat producing a 36% one-day move — forces a full rebuild, and rebuilds take weeks.
3. Check whether forward guidance also came in above consensus. That is what turns "one strong quarter" into "the whole forecast curve was wrong," and it is the condition under which target lag is largest.
4. Re-anchor on a cash multiple you compute yourself — run-rate EV/EBITDA or FCF yield off the new numbers — rather than on the target. The target is someone else's arithmetic on stale inputs.
5. Note the asymmetry: this reasoning applies at inflections, not to a stock that has simply drifted above a target on multiple expansion with flat fundamentals. If the numbers didn't move, the target isn't stale — the stock is just dearer.
6. Expect the targets to be revised up toward the price over the following weeks; that revision is itself a mild tailwind, not a reason to buy, but it warns against selling into the gap.

• The revised consensus target over the next few weeks — if it settles below the current price after analysts have rebuilt, the stale-model defence has expired and the valuation question is live again. Also watch for the first quarter in which guidance comes in at rather than above consensus, which ends the inflection.

2. Use a sequence of records, not a single print, as the evidence

1. Lay the last four to six quarters of the key metric side by side in order, and add the guided next quarter to the end of the row. One line of numbers replaces a page of argument.
2. Ask whether each step builds on the last. Consecutive records rule out the two most common bear explanations at once — a one-off comparison, and a pull-forward of demand.
3. Include the guided quarter deliberately: management's willingness to guide the sequence forward is the strongest available statement that they don't see a high-water mark.
4. Place the sequence inside the cycle. The same trajectory means opposite things early in a spending recovery and late in one — so state where the customer's budget cycle stands, and why.
5. Use the sequence to distinguish thesis confirmed from thesis accelerated. The second justifies pulling forward the timeline of a price target while leaving the target itself unchanged.

• The first quarter that fails to set a record, or guidance that flattens the sequence — that single break invalidates the "middle of the cycle" read far more cleanly than any valuation argument would.

3. Convert a big gain into a sizing decision, not a valuation decision

1. Separate two questions that a large gain tends to fuse: is the thesis still true? and is the position still the right size? A yes to the first with a no to the second calls for a trim, not an exit.
2. Trigger the review on portfolio weight, not on the percentage gain. The relevant fact is that the holding "has grown to an outsized share of total portfolio weight" — a number only the holder can compute.
3. Specify the trim as a fraction of the position (25-33%), so the instruction scales to any starting weight and leaves a core intact by construction.
4. Say explicitly that the remainder is a core holding, and why — otherwise a trim reads as a loss of confidence and holders exit the whole thing.
5. Where a previous note deferred the sizing question to a dated catalyst, answer it on schedule. Position-management guidance that never resolves is worthless.
6. Contrast with the exit discipline used on a name whose thesis is finished rather than working — dollar-cost-averaging out through a vertical "hockey stick" move. Trimming for weight and selling for froth are different actions with different triggers.

• The promised portfolio tables ("later this week") to see whether the trim is executed in the book itself and at what rating — the difference between advice to readers and action on the house position.

4. For a project business, read backlog and book-to-bill ahead of the earnings line

1. For companies that sell multi-year projects rather than units, identify the forward-revenue metrics: total backlog, its growth rate, and the trailing book-to-bill (new orders ÷ revenue recognised).
2. Read book-to-bill above 1.0× as the business winning work faster than it is consuming it. 1.4× means the pipeline is growing 40% faster than it is being burned — future revenue that is already contracted.
3. Rank these above the quarter's EPS, which for a project business is mostly a function of timing and mix. Explicitly say so, so a small EPS miss doesn't get read as a thesis break.
4. Corroborate with mix disclosures: which end-markets are growing, and are the structurally-favoured ones gaining share of revenue? A rising share from the theme you are underwriting is the cleanest confirmation.
5. Check the balance sheet against its own target — hitting a leverage goal early is management evidence independent of the demand story.
6. Book the weakness first. Stating the miss before the bull case is what makes the bull case credible.

• Book-to-bill slipping toward or below 1.0×, or backlog growth decelerating — either would mean the forward revenue is no longer compounding, and that is the metric that breaks this thesis, not an EPS miss. Also watch the AI-infrastructure share of revenue, which should keep climbing if the structural read is right.

5. Pair a positive view with a stated entry rule when your macro stance argues against chasing

1. When you like a business but distrust the tape, do not resolve the tension by weakening the rating. Keep the view positive and attach a price condition to the action.
2. Make the condition specific and small enough to be reachable — a 5% pullback is a normal week, not a market call. A condition that requires a crash is a polite refusal.
3. Name the reason for the restraint, and keep it separate from the company: "given our broad market concerns" is a portfolio-level input, not a downgrade of the business.
4. State what you expect while you wait, so the rule has a rationale beyond timidity — here, a stock that "catches its breath for a spell" before it can "make up for lost time."
5. Apply the mirror-image rule to the name that has already run hard: trim into strength there, wait for weakness here. The two instructions in a double update should be internally consistent about the same market view.

• A ~5% pullback to roughly $138 as the stated trigger to buy — and, on the other side, the stock running further without the pullback, which is the cost of the discipline and worth tracking as a scorecard on whether the rule adds value.