Actionable insights — Portfolio Update: May 2026

1. Keep intrinsic value and share price as two separate annual series

1. Each year, record two numbers for the portfolio: the growth in estimated intrinsic value, and the change in market price.
2. The difference is the change in valuation — record it as its own figure, in percentage points.
3. Cumulate the series so you can say how much cheaper or dearer the book has become over multiple years.
4. Compare that cumulative figure with the same computation for your benchmark, to separate what happened to you from what happened to markets.
5. Use the series to decide whether underperformance was fundamental (intrinsic value stalled) or valuation-driven (price fell faster).

• Your own intrinsic-value estimates drifting up because you want them to. This series is only honest if the method for estimating value is fixed and dated.
• The 2026 figure being an expectation, not a result. One of the two numbers in that year's comparison has not happened yet.

2. Estimate expected return with two inputs and no exit multiple

1. Compute the portfolio's weighted forward P/E, then invert it to get the earnings yield.
2. Add your expected earnings growth rate.
3. Treat the sum as the expected annual return, and note that it assumes the valuation does not change — which is the conservative assumption.
4. Track the number over time. It rises when prices fall, so it doubles as a valuation gauge for your own book.

• The 12% growth input doing all the work. Halve it and the expected return falls to 11.8%.
• A forward P/E computed on analyst estimates that are themselves being cut.

3. Use the cash metric the business is actually run on, not reported earnings

1. Ask which single number management uses internally and reports to shareholders as the measure of success.
2. For serial acquirers, reported earnings are depressed by amortisation of acquired intangibles — use free cash flow available to shareholders instead.
3. Check that metric's multi-year trend, and use its growth rate rather than EPS growth in your valuation.
4. Verify the definition has not changed between periods.

• Bespoke metrics that flatter. FCFA2S is a real and disciplined measure at Constellation, but a company-defined metric always needs its bridge to statutory cash flow checked.
• One quarter of 44% growth being annualised in your head.

4. When a price is cut, check whether volume more than compensates

1. Establish the size of the price cut and which part of the product line it applies to.
2. Find the volume data — patients, units, subscribers — for the period after the cut.
3. Ask whether the lower price opens a market that was previously inaccessible, or simply reduces revenue from the existing one.
4. Check the competitive consequence: does the cut also lock a rival out?

• Gross margin. A tenfold volume increase at a seventh of the price is only good if the unit economics survive it, and that number is not given here.
• Patent and exclusivity dates, which cap how long the volume belongs to you — semaglutide exclusivity is already gone in Canada and India.

5. Ask whether a competitor taking share is doing it profitably

1. When a holding loses share, identify who is gaining it and how — price, product, or distribution.
2. Compare returns on capital between the two. A loss-making challenger buying share with discounts is running a strategy with a time limit.
3. Distinguish that from a challenger with a genuinely better product, which is a permanent problem.
4. Check whether the incumbent's response is cost discipline and pipeline, or matching the discount.

• "A wave of new product launches" hiding inside the discounting story. That half of the Elanco threat is not a pricing problem and does not expire.
• An incumbent that eventually has to match the price to hold share, which converts a share problem into a margin problem.

6. Build a target price from three stated inputs and show the arithmetic

1. State the fair multiple you would apply, and why it is normal for this business.
2. State the earnings year you are underwriting to, and the earnings figure.
3. Multiply, and convert the result into both a percentage upside and an annualised return over the years to that date.
4. Publish all three inputs so any reader can substitute their own.

• Compounding two optimistic assumptions. The 142% needs both the multiple to double and the earnings to reach $9.
• The four-year gap. Nothing in the case explains what happens to the shares between the 11% companion-animal decline now and the 2027 product wave.

7. Distinguish a conviction statement from a capital commitment

1. When you (or a source) declare that conditions are exceptionally favourable, write down what action that would imply.
2. Check whether the action followed: new money deployed, position sizes increased, cash drawn down.
3. If the response was a reallocation rather than an addition, record it as a relative view, not a market call.
4. Hold the statement against the subsequent record.

• "Opportunity of a century" language recurring across issues — the same framing appears on 7 May. Repeated urgency is a style, not a signal.
• A fully invested portfolio having no way to swing heavily. Without cash, conviction can only be expressed by selling something else.