Foreign desks sold 351.9 billion rupiah on the session while retail bought 185.1 billion and the local institutional screen stayed positive at plus 46.5 billion. But the aggregate hides almost everything that happened today. Two names moved in opposite directions with a violence that the market-wide numbers cannot show.
ANTM — the checklist read
Four of the seven foreign desks bought, totalling 269.9 billion rupiah: UBS 95.3, CGS 75.2, J.P. Morgan 69.4 and Maybank 62.6. Retail sold 186.1 billion, the largest single line anywhere on the retail screen. Four desks arriving at similar size on the same day, against retail supply of almost identical magnitude, is the transfer pattern in its clearest form.
The local institutional screen reads minus 4.2 billion. Small, and negative.
That figure is the reason this reads as a strong signal rather than a complete one. Under a checklist that treats local participation as a requirement rather than a bonus, a negative number disqualifies regardless of how good the foreign side looks — and the foreign side here looks very good. The discipline of the rule is that it does not get to make exceptions for the cases it finds persuasive.
BBCA — the largest single line in the block
Maybank sold 384.8 billion rupiah of BBCA in one session. That is more than double any other line across all nine screens, and it drags the foreign block to minus 457.6 billion despite Macquarie buying 74.9 and CGS 37.2. Retail bought 277.6 billion of it.
Set against the rolling five-day picture, BBCA now shows minus 298.5 billion of foreign flow — the second-largest absolute figure on the board, and negative. A week ago the same name was among the most heavily accumulated stocks on the exchange. Positions of that size do not reverse quietly, and the fact that retail absorbed most of today’s supply is the part worth noticing.
BBRI — a signal that expired in one session
Yesterday BBRI showed six of seven desks buying, 270.7 billion net, and the local screen confirming at plus 21.5 billion. Today foreign remains a buyer at plus 104.6, but the composition has changed underneath it. CGS still carries 98.6 and UBS 39.6, while J.P. Morgan flipped from plus 31.8 to minus 38.0 and CLSA from plus 47.3 to minus 25.1. Six desks buying became three.
The local screen went from plus 21.5 to minus 49.2.
This is the same shape the flow took on 21 and 31 August, when foreign bids of 469.7 and 633.7 billion were not joined by the local desks and gave way within three sessions. A broker-flow read is only as current as the desks behind it. Yesterday’s reading was accurate for yesterday. It is not a description of today.
TINS and AADI — the near versions
TINS shows plus 81.5 billion across three buying desks with the local screen confirming at plus 11.2 and retail selling 41.1. AADI shows plus 115.4 billion across five desks, all five buying, with retail selling 33.2 — but the local screen sold 23.5 billion into it.
The contrast between the two is the whole lesson in miniature. Both had foreign demand. Only one had the domestic desks alongside it.
The screens truncate at roughly twenty-two lines per side per desk, and what falls below is absent rather than zero. BBCA, ASII, KLBF and BRPT had no local institutional line at all today, so their local flow is unknown, not neutral.
There is a subtler limit visible in ANTM. Four desks bought, but three did not — CLSA sold 17.1, Macquarie 12.1 and Henan 3.4. The net of 269.9 is what remains after the disagreement, and a single figure describes the balance of opinion without revealing that opinion was divided.
And as always, flow is not inventory. A desk selling today may be trimming a large holding or establishing a new short, and the daily summary cannot distinguish them.
A signal has an expiry date, and it is usually shorter than it feels.
BBRI cleared every condition yesterday on evidence that was real and verifiable. One session later, two of the desks that made the case had reversed and the local screen had gone the other way. Nothing about yesterday’s reading was wrong. It simply stopped being true, and it stopped being true faster than a reader relying on it would comfortably expect.
A reader watching this tape would note that the useful question is never only what the flow showed. It is also how recently, and whether the participants who produced the reading are still behind it.
Analisis ini bersifat edukasi, bukan nasihat investasi. Label “ACCUMULATION”, “WATCH”, dan “DISTRIBUTION” menggambarkan pembacaan data aliran broker, bukan perintah transaksi. Segala keputusan investasi beserta risikonya merupakan tanggung jawab masing-masing pengguna. Lakukan riset mandiri (DYOR).