The Dilution Tax

Issue #9-1 closes a gap left by the valuation frameworks in Issue #9. Those frameworks — EPV, two-stage DCF, residual income — all produce per-share figures, but a per-share figure is only as reliable as the share count used to compute it. Companies change that denominator constantly, and the direction matters as much as the magnitude.

The issue opens with a real-time example. On April 28, 2026, CATL announced a HK$39.2 billion private placement: 62.4 million shares at HK$628.20, a 7% discount to the prior close of HK$675.50. The stock fell 9.2% to HK$613.50 the same day — not because the business had deteriorated (Q1 2026 profit was up 49% year on year) but because existing shareholders woke to find their per-share claim diluted by an issuance they had no part in. The example is deliberate: dilution risk is not a story about weak or struggling companies. It operates across the quality spectrum.

Three issuance mechanisms carry distinct implications. Share buybacks, executed below fair value, reduce the denominator and raise earnings per share mechanically — the classic deployment of excess capital that directly benefits continuing shareholders. Rights issues are structurally neutral: all shareholders receive the right to participate in proportion to their current holdings, so no ownership percentage is transferred to outsiders. The dilution only bites shareholders who choose not to participate. Private placements are the mechanism to watch most carefully. Selected institutional investors receive new shares at a discount, transferring a portion of the company’s equity value from the existing shareholder base to the new entrants. The dilution is immediate, involuntary, and invisible unless the investor is tracking the share count.

The Dilution Dashboard introduced in the accompanying Colab notebook operationalises this analysis across a five-year window. It computes the compound annual growth rate of shares outstanding, decomposes net cash flow from financing activities into gross issuance and gross buyback, and assigns each company a traffic-light signal: NET BUYER (share count declining, CAGR below zero), STABLE (annual dilution at or below 4%), or SERIAL DILUTER (annual dilution above 4%). The threshold of 4% is not arbitrary — it reflects the level at which dilution begins to meaningfully offset the EPV per-share calculation on a three-year forward basis.

The connection to earlier issues in the fundamental-analysis track is explicit. Issue #8 introduced the Piotroski F-Score, which includes a NEWSHARES signal that penalises any net equity issuance in the prior year. The Dilution Dashboard extends that binary signal into a quantified adjustment: the historical dilution rate is applied forward to project the share count three years out, and the EPV computed in Issue #9 is restated on a per-share basis using that adjusted denominator. A company that looks cheap at today’s share count may look fairly valued — or worse — once the expected issuance is priced in.

The academic grounding is robust. Pontiff and Woodgate (2008) showed that net share issuance carries statistical significance in predicting future cross-sectional returns that exceeds the size premium, the book-to-market premium, and the momentum factor. Fama and French (2008) confirmed it as one of three anomalies that survive across all market-cap categories. McLean, Pontiff and Watanabe (2009) extended the finding internationally. The dilution screen is not a novel insight — it is one of the most empirically validated signals in the cross-sectional return literature, and one that most retail investors ignore entirely.

The framework does not replace valuation or quality analysis. It completes them. A high Piotroski score and a below-EPV price are necessary but not sufficient conditions for a long position. The share count trend is the third gate. Passing all three — quality, valuation, and dilution — is a significantly higher bar than the market’s typical filter, and that selectivity is precisely the point.