Sheng Siong: I priced the margin and missed the volume
The underweight rested on same-store sales fading. Same-store sales came in at 3.6% against my 2.1%. Scored wrong.
Sample post-mortem with illustrative figures. Two paragraphs on the archetype, one on the country, one lesson.
What I said
The underweight opened on 30 July at S$1.72 rested on same-store sales fading as HDB tender competition rose, with a kill condition of same-store sales above 3% for two halves.
What happened
Same-store sales came in at 3.6% against my 2.1%. The tender pipeline I was tracking was the wrong series; new-store cannibalisation did not show up because the new stores were in estates without an incumbent. The stock rose and the call is scored wrong on the reason, not only on the price.
What the archetype taught
For a volume-times-price compounder, model the store cohort, not the aggregate. Aggregate same-store sales hides where the growth is coming from, and the cohort split was available in the results deck.
What the country taught
Singapore grocery is a two-player market in which HDB tender outcomes are public. I should have built the tender calendar into the model rather than treating competition as a qualitative risk.
Lesson for the one-page framework
Cohort before aggregate. Any volume driver that can be split by vintage should be.