The Holdings-Sector Paradox: Top of the Cap, Bottom of the Grade

SM Is a Poison Apple

SM Investments Corporation is the largest issuer on the Philippine Stock Exchange by market capitalization. Its Apex G-Score archetype is Poison Apple — T 70.0, R 86.7, B 45.0 — strong on transparency and conflict-of-interest controls, weak on board substance. Ayala Corporation, ranked fourth by market cap, reads as Hidden Gem at total 69. JG Summit Holdings, seventh, and Aboitiz Equity Ventures, ninth, read as Chameleon[B-weak]. Among the ten largest issuers on the PSE by market capitalization, four are Holdings-sector parent vehicles^[1]. Not one reads as Celestial.

The pattern compresses further at the sector level. The PSE's Holding Firms sector, which contains thirty-one issuers including all seven family-conglomerate parent vehicles in the Apex framework's mapped cohort, carries a D-grade rate of 83.9%. That is the highest D-rate of any multi-issuer sector on the exchange. The sector that anchors the top of the Philippine market by capitalization is also the sector that anchors the bottom of the grade distribution. The two facts coexist in the same row of the framework's data.

Where the Sectors Sit

The full sector grade distribution, sorted by D-rate descending, is the following^[2]:

Holdings carries the highest D-rate of any multi-issuer sector — 83.9%, against the universe average of 76.0%. Services follows at 82.3%, Property at 76.5%, Industrial at 71.6%, Financials at 71.0%, Mining and Oil at 70.8%, and the SME Board at 66.7%.

The sector axis means complete the picture. The B-axis is the weakest axis in every PSE sector — Holdings averages 39.4 against universe 43.7, the lowest sector-level B-axis on the exchange. Property reads highest on T-axis (63.9), Financials and Mining and Oil read lower on R-axis (55.2 and 57.5). The structural pattern documented in Note 1 — uneven-axis profiles dominating the listed universe — lands on the same axis in every sector slice the framework reads. Holdings is not unique in being B-weak. It is unique in being the most B-weak.

Banks within the Financials sector read differently from Holdings on every measure. The Banks subgroup carries an 11.8% Kill Switch rate — the highest sector-level KS rate in the universe — but a B-axis mean of 49.1, materially better than Holdings' 39.4. Holdings carries a 6.5% KS rate but the lowest B-axis mean and a Poison Apple share of 29.0% against the universe's 17.3%. The two flagship sectors of Philippine governance read in opposite directions: Banks fail at higher rates with a less-weak board axis; Holdings carries fewer outright failures but a board axis that pulls every parent vehicle below the Celestial threshold^[3].

Top of the Cap, Bottom of the Grade

The Holdings paradox is not produced by a small number of mid-cap issuers pulling the sector mean down. It is produced by the largest companies on the exchange.

Among the thirty largest PSE issuers by market capitalization, seven sit in the Holdings sector: SM Investments, Ayala Corporation, JG Summit Holdings, Aboitiz Equity Ventures, San Miguel Corporation, GT Capital Holdings, and LT Group. Holdings has the largest Top-30 sector representation of any sector classification on the exchange. Its D-rate is also the highest. The cross-tabulation isolates the paradox in a single row:

Holdings is the only sector on the PSE where the highest concentration of top-cap issuers and the highest D-rate co-occur. Industrial matches the Top-30 count at seven, but its D-rate is twelve points lower and its Top-30 composition includes operating businesses — Jollibee Foods, AboitizPower, Manila Electric Company — rather than parent vehicles. Services carries an 82.3% D-rate but only six of thirty top-cap issuers; Property carries 76.5% with five. The Holdings sector is structurally the parent-vehicle layer of the family conglomerates that anchor the PSE — and that layer reads, on the framework's measure, weaker than the operating businesses it owns.

Seven Parents, Four Archetypes

The seven family-conglomerate parent vehicles within the Holdings sector span four distinct archetypes despite all sitting in the same sector classification:

The seven parent vehicles split across four archetypes. Hidden Gem: Ayala Corporation, alone. Poison Apple: SM Investments and Lopez Holdings. Chameleon[B-weak]: Aboitiz Equity Ventures, House of Investments, and JG Summit Holdings. Time Bomb: San Miguel Corporation.

Among the seven, only AC and SM clear C-grade; the remaining five carry D-grade. AC's Hidden Gem reading reflects the strongest combined T-and-R profile across the seven, with a B-axis at 62.5 — the highest B-axis score among all family parent vehicles. SM's Poison Apple reading reflects an R-axis of 86.7 — the highest of the seven — pulled down by a B-axis of 45.0. SMC's Time Bomb classification is mechanical: all three normalized axes sit below fifty (T 43.3, B 40.0, R 40.0), the structural floor that defines the archetype. SMC carries an unqualified audit opinion on the most recent fiscal year. The Time Bomb reading is a structural-archetype signal, not a regulatory distress flag.

The single observation worth carrying forward is that Holdings is not monolithic. The framework reads the seven family-conglomerate parents as carrying fundamentally different governance signatures despite a shared market-capitalization tier and a shared sector classification. Cell sizes are small enough that single-family rates produce wide confidence intervals on any per-group claim — the seven-family cohort is too small for percentage-based statements about the cohort as a whole. The four-archetype dispersion is itself the finding^[4].

Two Architectures

The seven family parents do not relate to their operating subsidiaries the same way. The framework reads two distinct family-architecture patterns in the data.

In the first architecture — represented by Ayala Corporation and SM Investments — the parent vehicle reads stronger than its operating subsidiaries on the framework's composite. AC's total of 69 places it above the average score of its seven listed subsidiaries (Ayala Land, BPI, Globe, ACEN, AyalaLand Logistics, AREIT, IMI), which average around 57. SM's total of 65 places it above the mean of SM Prime, BDO, and Converge (around 56). The flagship parent vehicle, in this architecture, carries the cleanest governance reading in the group.

In the second architecture — represented by JG Summit, San Miguel, and House of Investments — the parent vehicle reads weaker than its operating subsidiaries. JGS at 46 against subsidiaries Universal Robina, Robinsons Land, and Cebu Pacific averaging 53. SMC at 41 against San Miguel Food and Beverage and Ginebra San Miguel averaging 48.5. HI at 48 against RCBC at 54. The parent vehicle, in this architecture, absorbs the structural weakness — pyramid cross-listing, intra-group RPT, layered control — while the operating subsidiaries are kept in better governance condition.

The Aboitiz and Lopez parents read at roughly the same level as their subsidiaries — neither cleaner nor materially weaker.

The cell-size constraint is severe. Only Ayala carries enough listed subsidiaries (n=7) to support a per-family parent-vs-subsidiary statistical claim. The other six families produce one to three subsidiaries each, which generates wide confidence intervals on any single-family roll-up. The pattern is presented as a typology rather than as a distribution: two architectures, observable in the data, with single-family claims qualified throughout. Note 3 returns to the MERMAC pyramid and the cross-listing reading at depth.

Why the B-Axis Sinks

The Holdings B-axis mean of 39.4 is the lowest sector-level B-axis score on the PSE. Two indicator-level signals carry the floor.

The independent-director count and 9-year tenure recycling check sits at the bottom of the B-axis distribution within the Holdings cohort — the same indicator the framework reads as the weakest B-axis component universe-wide. The chair-CEO separation indicator, conditional on family officers in operating roles, places more than forty percent of Holdings-sector issuers at the floor band. These two indicators together explain most of the sector's B-axis weakness; the remaining four B-axis components sit in the mid-band but do not collapse to floor with the same frequency.

The R-axis tells a more nuanced story. The Holdings R-axis mean of 58.7 is actually 1.4 points above the universe average. Most R-axis components in the Holdings sector read at or above universe levels — private placement discount, related-party lending, and audit opinion severity all sit at or above the universe baseline. The R-axis penalty in Holdings is concentrated in one indicator: pyramid cross-listing. The largest cross-listing meshes — Ayala's eight listed entities, San Miguel's seven — drive the Holdings R-axis below where it would otherwise sit, but the rest of the R-axis profile carries Holdings above the universe mean. Note 3 takes up the cross-listing reading at depth.

Two of the thirty-one Holdings-sector issuers carry the framework's Kill Switch override. Filinvest Development Corporation, the more visible of the two, reads as KS-7 structural: ultimate beneficial ownership concentration at or above eighty-five percent — the framework's threshold for the structural KS classification. KS-7 is a control-concentration reading, not a regulatory distress signal — Filinvest carries no current-year audit issue and operates through the standard PSE disclosure cycle. The second KS-flagged Holdings firm reads on the same KS-7 structural condition. The 6.5% Holdings-sector KS rate sits below the 7.1% universe rate. Holdings concentrates Poison Apple, not KS, against the universe baseline.

A Regional Cluster, Different Architecture

The Korean listed universe carries a parallel pattern. Non-financial holding companies in Korea over-represent in the Poison Apple archetype — the same T-strong, B-weak signature the Apex framework reads in the PSE Holdings sector^[5]. The Philippine Holdings sector reads at +11.7 percentage points over the universe Poison Apple share; Korean non-financial holdings show a directionally similar over-representation against the Korean universe baseline.

The two markets cluster on the holding-company-as-Poison-Apple signature. They diverge on the family architecture that produces it. Korean holdings emerge from circular shareholding structures, restricted by Fair Trade Act amendments dating to 2014^[6]. Philippine holdings emerge from constitutional-ownership-cap-driven layered structures, including Philippine Depositary Receipt instruments and voting trusts that work around the 60/40 foreign-ownership ceiling. The mechanical T-strong, B-weak outcome is the same. The legal-structural pathway is different. Note 6 takes up the Korea-Philippines comparison in depth.

The Holdings sector reading on the PSE — highest market-cap concentration, highest D-rate, lowest B-axis mean, four archetypes across seven parents, two distinct family architectures — captures the central paradox of Philippine governance at the issuer level. The largest Philippine companies are also, on the framework's measure, the companies whose governance variation is widest and whose structural weakness is most concentrated. The sector that anchors the Philippine market is the sector where the framework's reading carries the most disagreement.

The Apex G-Score framework currently covers 283 Philippine listed companies under the v1.0 Philippines calibration. Distribution figures reflect the FY2024 production reference. Top-30 market-capitalization references in this Note are anchored to public market-capitalization data as of FY2024 (PSE EDGE, Bloomberg) rather than a system-derived ranking; rank order shifts quarter-to-quarter while the top-tier sector composition is stable.