Singapore — End-of-Day Verdict
31st cycle: the acquire side reopens after two empty cycles as S68 SGX prints the board's only bullish actionable row — STRONG BUY - ADD cv99 at BuyRank #1 on a monotonic → conviction ramp — but it is 2 ATR extended, so SG-6's cv≥95 immediate-tranche carve-out is rejected on three grounds and the ADD is queued at 24.27; and O39 becomes the THIRD live decision blocked on the same unreconciled holdings file, leaving the dispose side structurally empty
Recommendations
1 to acquire · 0 to dispose
| Ticker | Name | Close | Action | Action Price Where to act — a buy-limit on a pullback or a buy-stop above resistance, not the current price. Sanity-check vs Close: on a held name, an action price stranded above the close means the queued level was set when the stock was higher — stale, awaiting reprice. | Stop The risk-defined exit, set below the entry for a long (~1.2–1.8×ATR). A held long whose Stop sits at/above the Close is already below its stop — a stale box awaiting reprice. | 1wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | 4wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | 12wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | Rationale |
|---|---|---|---|---|---|---|---|---|---|
| S68 | Singapore Exchange | 25.24 | ADD — QUEUED @24.27 | 24.27 | 23.29 | 25.21 (+3.9%) | 24.98 (+2.9%) | 24.88 (+2.5%) | Singapore Exchange is the strongest setup on the board: the settled close prints STRONG BUY - ADD at conviction 99 on a PullbackBuy_VWAP signal at Buy_Rank #1, the fourth straight session of strengthening conviction to on a name up roughly 36% year-to-date as Asian capital flows lift SGX's and derivatives volumes. The catch is price: the close sits about 2 ATR above the model's buy level and an independent long-run cash-flow appraisal, like our own residual-income read, puts the shares above intrinsic value - so we bid for the pullback rather than chase the print. OPERATIONAL: extended at Entry_Distance_ATR -1.97, so the SG-6 immediate-tranche carve-out does NOT apply (its own exclusion keeps the strict no-chase for extended names) and the action-price provenance rule pins the limit to the live Entry_Price 24.27 +/- 0.25 ATR (24.15-24.39), which also rules out a shallower GTC. The estimator independently rejects the chase: Est_1W 25.21, Est_4W 24.98 and Est_12W 24.88 are ALL below the 25.24 close, so SG-6's premise that the name runs away is falsified here. Sized : raw ADD 114.38% x 3000 =, x0.75 size modifier (MoS -1.82 in the -3 to -1 band) =, cut by the declared Sdry to / Smax_position_pct 60% was not the binding cap (headroom). Counterweight recorded: the no_chase_queue counterfactual is n=33, median withheld +5.45%/1wk and +11.18%/4wk, verdict COST - this queue may not fill and that has been the book's most expensive habit. PullbackBuy_VWAP|buy is 0/4 on 1wk targets, below the n>=5 floor, so NO bias was applied. No S68 dividend in the 10-day Event Horizon, so the 24.27 limit cannot be hit by a mechanical ex-div drop. Est_Model_Ver is flagged stale (sg/20260703T083907Z). Withdraw if conviction falls below 65 or unfilled by 2026-08-19. · news |
| O39 | OCBC Bank | 31.79 | HOLD | 31.67 | 30.22 | 31.64 (-0.1%) | 31.29 (-1.2%) | 31.95 (+0.9%) | The profit-take case on OCBC is still valid on the merits - the position is up about 22.6% on a real, momentum is stretched at RSI 78, and the twelve-week view has gone flat - but the same trim has now been published three sessions running without ever being executed, so this is an execution problem, not a decision problem. No new card is issued until the holdings file reflects reality. OPERATIONAL: SG-8 fires exactly as the 08-13 entry promised. Published 08-11 (50%), 08-12 (30%), 08-13 (30%) with input/sgportfolio.json byte-identical since 2026-07-15 - 21st consecutive static cycle. Decision NOT reversed and size NOT re-escalated. For the record the ADR-0012 fire still holds on 2 ticks: (a) RSI_DIV at RSI 77.87, (c) dead 12wk 31.95 < 1.02 x 31.79 = 32.43; tick (b) does NOT fire, conviction is only -15. Armed on the one REAL Avg_Cost 25.9325 (SG-7): Gain_ATR 8.05, T12_Progress 99.5, +22.59%. Trail RAISED 30.05 -> 30.22 (live model stop). SCHEDULED REBASE 30.22 -> 29.75 on 08-17 (47c interim ex-div, confirmed). Unblocks the moment the holdings file is synced. · news |
| B61 | BT SEMBAWANG EST | 4.50 | HOLD | — | 4.62⚠ | 4.49 | 4.50 | 4.56 | Bukit Sembawang has closed below its risk line for a sixth straight session, but the exit that was decided a week ago has still never been executed, so the book re-states the blocking flag rather than re-deciding the trade. On the evidence the exit was never well-founded anyway: the twelve-week view sits above the current price and the shares screen as undervalued. OPERATIONAL: SG-8 cycle 2. Close 4.50 vs the 4.62 stop (-2.60%) on a Sposition. Est_12W 4.56 is +1.3% ABOVE close and the NoTrade|sell cell is ACTIONABLE at -5 (n=7, hit_1w 0.25, hit_4w 0.00), a direct haircut on this card. Sdividend pays 08-15. Stop held at 4.62. Unblocks on holdings sync. |
| HMN | CAPITALAND and ASCOTT STAPLE U | 0.86 | HOLD | 0.85 | 0.88⚠ | 0.86 (+1.2%) | 0.86 (+1.2%) | 0.88 (+3.5%) | Eighth consecutive close below the stop on a position worth barely S - the decision has been made repeatedly and never executed, so it stays a blocking sync flag rather than a fresh card. OPERATIONAL: SG-8 cycle 6. Close 0.855 vs the 0.88 stop (-2.84%), = SSignal weakened AVOID cv50 -> cv40 and Est_12W 0.88 is +2.9% above close. Stop held at 0.88. Unblocks on holdings sync. |
| A17U | CapitaLand Ascendas REIT | 2.44 | HOLD | 2.44 | 2.42 | 2.44 (+0.0%) | 2.46 (+0.8%) | 2.45 (+0.4%) | Ascendas REIT throws the loudest sell label on the board tonight, but it fails the book's own confirmation test: the price is still above its risk line and the twelve-week view is positive. SG sell labels reverse more often than not, so the label alone is not a reason to sell a REIT that has not actually broken down. Holding. OPERATIONAL: STRONG SELL - REDUCE cv61 at Sell_Rank #1, but SG-2 requires stop-break AND negative 12wk - close 2.44 vs the rebased 2.42 stop is NO break, and Est_12W 2.45 is +0.4% POSITIVE. RallySell_BearTrend|sell is ACTIONABLE at -3 (n=9, hit_1w 0.333), taking adjusted conviction to 58, below the 65 floor. Valuation is deeply negative (MoS -7.685; SG-3 says read the magnitude, not the REIT band) and WOULD amplify - but valuation never originates a sell (CLAUDE.md section 9), and there is nothing here to amplify. Size would have been trivial anyway: 80.07% x 400 = = SStop held at 2.42. Re-arm: a decisive close below 2.42 on a NEGATIVE Est_12W. · news |
| J69U | FRASERS CPT TR | 2.20 | HOLD | 2.20 | 2.18 | 2.20 (+0.0%) | 2.22 (+0.9%) | 2.26 (+2.7%) | Frasers Centrepoint prints a strong sell label for the eleventh time and fails the same confirmation test every time: it is still two cents above its risk line and the twelve-week view is the most positive of any flagged name. A label that has cried wolf eleven times on a trust that keeps holding its level is not an exit case. OPERATIONAL: STRONG SELL - REDUCE cv61 at Sell_Rank #2. SG-2 fails for the 11th time - close 2.20 vs the 2.18 stop is NO break, Est_12W 2.26 is +2.7% positive. RallySell_BearTrend|sell bias -3 puts adjusted conviction at 58, below the 65 floor. No valuation row (NA), so no amplify path exists. 80.07% x 28,000 would have been = S - the largest sell the board offered and correctly declined. Stop held at 2.18. Re-arm: close below 2.18 on a NEGATIVE Est_12W. · news |
| U11 | United Overseas Bank | 41.80 | HOLD | — | 42.15⚠ | 41.81 | 41.68 | 41.84 | UOB is the book's largest position and has now closed below its risk line three sessions running, but the forward view refuses to turn negative - it is positive by a tenth of a percent - and the 88-cent interim dividend on Monday mechanically resets the risk line below the current price. Holding through the dividend rather than selling into it. OPERATIONAL: 3rd consecutive break, close 41.80 vs the 42.15 stop (-0.83%) on = SL2/SG-2 FAIL again - Est_12W 41.84 is +0.10%, positive by a hair. Signal collapsed AVOID cv41 -> NO TRADE cv0 and the NoTrade|sell cell is ACTIONABLE at -5 (n=7, hit_1w 0.25, hit_4w 0.00), so a NoTrade-originated sell here carries a full -5 haircut. SCHEDULED REBASE 42.15 -> 41.27 on 08-17 (88c interim ex-div, independently confirmed) - this clears the breach mechanically. WATCH TRIGGER stays armed and is one limb from firing: a close below 41.27 after the rebase on a NEGATIVE Est_12W makes U11 the immediate lead dispose. · news |
| D05 | DBS Group Holdings | 75.53 | HOLD | 75.53 | 73.04 | 75.19 (-0.5%) | 74.94 (-0.8%) | 77.89 (+3.1%) | DBS went ex-dividend today for 81 cents and the tape absorbed it cleanly - the price is essentially unchanged once the payout is stripped out, and the twelve-week view remains the most positive of the three banks. No action; the risk line simply moves with the ex-dividend. OPERATIONAL: 81c (66c ordinary + 15c capital return) went ex 08-14 as confirmed; close 76.50 -> 75.53 is -0.16 net of the payout. The live model stop 73.04 SUPERSEDES the manually scheduled 72.89 rebase and is set as the operative stop - no phantom-breach artifact this time because the tape applied the ex-div itself. Label is HOLD cv51, NOT in the buy pool; conviction faded. SG-4 governs bank cores - no chase. Earliest legitimate re-arm is a non-extended BUY - ADD reprint at cv>=65. · news |
| N2IU | Mapletree PanAsia Com Tr | 1.28 | HOLD | 1.28 | 1.27 | 1.28 (+0.0%) | 1.29 (+0.8%) | 1.29 (+0.8%) | Mapletree PanAsia has climbed back above the risk line it was sitting exactly on last night, and its twelve-week view is positive. The AVOID label is not backed by a breakdown. Holding. OPERATIONAL: close 1.28 vs the 1.27 stop - RECLAIMED after touching exactly at the 08-13 settle. AVOID cv62 but FailedBreakout_Down_20D|sell is ACTIONABLE at 0 (n=73, hit_1w 0.507) - a coin flip, per SG-5. Est_12W 1.29 positive. Stop held at 1.27. |
| BUOU | FRASER L&C TRUST | 0.96 | HOLD | — | 0.96 | 0.96 | 0.96 | 0.99 | Frasers Logistics is half a cent above its risk line with no signal at all, and is the next name in the book likely to break. Valuation is poor on the magnitude read, but a poor valuation has never been a reason to sell on its own. Watching, not acting. OPERATIONAL: close 0.965 vs the 0.96 stop. NO TRADE cv0 means there is no signal to originate a sell, and MoS -6.033 (SG-3: read magnitude, ignore the FAIR_VALUE band) can only AMPLIFY a signal-based case that does not exist. = Sthe book's second-largest position. Stop held at 0.96 - a decisive break there with a negative Est_12W is an immediate L2 dispose candidate. |
| C07 | Jardine C&C | 27.58 | HOLD | — | 26.38 | 27.85 | 28.54 | 28.55 | Jardine Cycle & Carriage remains the single most attractive name in the book on intrinsic value - the only holding that is undervalued, passes the quality screen and offers a genuine discount - but it prints no signal whatsoever, and cheapness alone is not an entry. Waiting for the technical to catch up to the valuation. OPERATIONAL: MoS +0.628, Quality_Pass True, val_adj +10 (the maximum), DCF 65.46 vs close 27.58. NO TRADE cv0, so adjusted conviction is 10 - far below the 65 floor. Est_12W 28.55 is +3.5%. Re-arm on any BUY-family reprint at cv>=55. Note the 08-20 Jardine Matheson ex-dividend is a correlated drag on JC&C (Event Horizon, low confidence). |
| S58 | SATS | 4.69 | WATCH | 4.69 | 4.65 | 4.68 (-0.2%) | 4.68 (-0.2%) | 5.02 (+7.0%) | SATS still has the best forward shape of any name the book does not own, but the signal has faded to a WATCH and the valuation layer now marks it well above intrinsic value - not enough to justify committing the last of the. OPERATIONAL: WATCH cv44, Est_12W 5.02 = +6.8%, the strongest unheld 12wk on the board. Regenerated MoS -2.346 gives val_adj -10, so adjusted conviction is 34, well below the 65 floor. Re-arm on a BUY-family reprint at cv>=65; the 08-05 buy-stop queue was withdrawn on 08-06 and is not reopened. |
| VC2 | Olam Group Ltd. | 1.20 | WATCH | — | — | 1.19 | 1.20 | 1.25 | Olam reported first-half results today and screens as undervalued with a passing quality check, but the signal engine gives it nothing to act on and the book does not chase a results-day print. Watching for a clean setup. OPERATIONAL: 1H 2026 results landed 08-14 (Event Horizon, two-sided +/-1.0%, low confidence). MoS +0.479 -> val_adj +10; NO TRADE cv0 -> adjusted conviction 10, far below the floor. L6 - no post-results gap-chase. Est_12W 1.25 = +4.2%. Re-arm on a BUY-family print at cv>=55. |
| S63 | ST Engineering | 10.96 | HOLD | — | 9.58 | 10.96 | 11.02 | 10.79 | ST Engineering is well above its risk line and quiet after results - no signal, no action. OPERATIONAL: NO TRADE cv0 on 'Liquidity gate failed'. Close 10.96 vs the 9.58 stop, comfortable. Est_12W 10.79 is -1.6%, mildly negative, but there is no stop break so L2 does not engage. MoS -9.067 amplify-only. Stop held at 9.58. |
| 9CI | CapitaLand Investment | 2.74 | HOLD | — | 2.62 | 2.73 | 2.76 | 2.85 | CapitaLand Investment holds its post-results gain with the strongest twelve-week view among the book's quiet holdings, but prints no tradable signal. Holding. OPERATIONAL: NO TRADE cv0; the 08-13 intraday BUY - ADD cv70 decayed to HOLD at the settle and did not reprint today - L4/L6, no gap-chase. Close 2.74 vs the 2.62 stop. Est_12W 2.85 = +4.0%. Watch for a non-extended BUY - ADD reprint toward 2.65. |
Outcome & track record
Accuracy at the time of this record.
116 Graded decisions Past recommendations old enough to score against what actually happened (target hit / stopped out / closed). This is the evidence base — the bigger the number, the more history behind the calls. Click to inspect every one. See the decisions →
0.229 1-week calibration How much of the eventual 4-week move typically shows up in the first week (0 = almost none, 1 = most of it). Use it to read the 1-week estimates: a low number means prices drift toward the target slowly, so don't over-read a soft week-1.
50 Published records How many dated end-of-day verdicts exist for this book — the length of the public track record you can browse in the Archive below.