This website is for educational purposes only and does not constitute investment advice. Always do your own research and assess your own risk tolerance before making investment decisions.

SG End-of-Day Verdict 17 Aug 2026, 17:00:00 GMT+8

SG Off-the-Radar — End-of-Day Verdict

NO ACQUISITIONS + 3 DISPOSE + 1 QUEUE WITHDRAWAL. No name in the book clears the 65 conviction floor as a buy, so the ACQUIRE bucket is an affirmative NO ACTION and the ANTICIPATE slot stays empty. The 08-14 S68 GTC @Sis WITHDRAWN: it never filled while SGX ran to S (+5.1% above the limit) and conviction collapsed, tripping the ADR-0012 auto-withdraw two sessions early; adj 49 after val_adj -9 leaves it 16 points under the floor. SGX's business is fine (H1 net revenue +7.6%, dividend raised to 11c) - this is a price and signal-quality rejection, and it re-arms on a fresh PullbackBuy_VWAP at raw cv >= 74. Closest miss is S63 ST Engineering at adj 64, one point short, and separately ineligible for the anticipate slot with only required ANTICIPATE prints - the missing 08-14 print was erased by the liquidity-gate misfire, but a print that does not exist cannot be inferred. THREE TRIMS, all held, all 30%: J36 JARDINE MATHESON - @Son §9 sustained non-buy now at N=6, with the 12-week view falling every session to -12.9% and the close 3% below its 200-day average; the full-exit clause did NOT fire (close 62.18 vs the 61.76 line) so it stays live on the remaining, and UNDERVALUED MoS +0.38 holds the tier at the 30% floor rather than escalating. O39 OCBC - @Son ADR-0012, armed at Gain_ATR 4.27 (+10.2%) with a THIRD consecutive momentum divergence at a rising RSI 78.9 and price 39.7% above EMA200; read the 24c dip correctly - the 47c dividend went ex TODAY, so like-for-like the shares ROSE and conviction rose : this is a trim on extension, not deterioration. G07 GREAT EASTERN - @Son ADR-0012, the most armed position in the book at Gain_ATR 8.34 (+24.5%) with a fresh divergence at RSI 81.9, selling into a dividend run ahead of the 08-19 ex-date; the same row simultaneously prints BUY - ADD at adj 68, and the add is suppressed by SGO-5 on both of its conditions at once (RSI > 75 and entry above the close) - the mandatory profit-take wins the collision. Net realised +Son Sof proceeds. HOLDS: D05 7,700 with the trail HELD at 74.59 because the printed stop came in lower and a trail never steps down - conviction has decayed -> 46 with divergence at RSI 72, which would be tick (b) but D05 is NOT armed, so it is a watch item; S58 60,200 with the trail raised to 4.61 as conviction recovered. Zero trail erosion book-wide. THE REAL FINDING IS NOT RISK, IT IS IDLENESS: the book sits at 50.3% cash and these trims take it to 51.5%, against a no_chase_queue counterfactual reading [cost] at a median withheld +5.24% 1wk / +11.04% 4wk across n=24 - two independent measurements of the same defect, entries priced too far below market and capital that never. Systemic: the liquidity gate is inverted or spike-bounded, not a low-volume floor - it blocked the two largest turnovers in the book on 08-14 and a 1.68m-share name today while passing a 111k-share name, and it cost the anticipate slot by erasing S63's second print.

Recommendations

0 to acquire · 3 to dispose

TickerNameCloseActionAction 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
J36Jardine Matheson62.18SELL 30%62.1861.76n/an/a54.16 (-12.9%)Jardine Matheson has now printed a non-buy signal for six straight sessions and its 12-week outlook has fallen every single session to 12.9% below the current price, with the shares trading below every moving average including the 200-day. We are cutting a further 30% of the position while the valuation still looks cheap on paper - a cheap price is not a reason to hold a falling trend, but it is a reason to trim the minimum rather than exit outright. Origination: CLAUDE.md §9 sustained non-buy, N=6 (NO TRADE x2 then AVOID x4), Target_12Week monotonic 59.40->54.16 (-12.9%), close 62.18 is 3.0% below EMA200 65.05. The 08-14 full-exit clause did NOT fire - it required a decisive close below 61.76 and the settled close is 62.18 (L4), so the ladder continues at the 30% floor and the full-exit clause stays live on the remaining. PROFIT_ARMED is a false positive for the 7th session (loss, Gain_ATR -0.86; degenerate T12_Progress 114.81) - this is NOT an ADR-0012 fire. UNDERVALUED MoS +0.38 amplifies DOWN, holding the tier at 30%. 30% x 900 = 270 -> floor to 100-lot =. Realised -S (cost 63.55). 1wk/4wk n/a - the bearish signal prints no long Target_Price, stated not synthesized. MANDATORY: trail steps 61.76 -> 61.11 on 08-20 (65c ex-div adjustment, NOT a lowering); printed Stop_Price 65.35 is the short-side stop, not applicable to a long. Ladder ->. · news
O39OCBC Bank31.55SELL 30%31.5530.0932.10 (+1.7%)35.56 (+12.7%)41.66 (+32.0%)OCBC is up 10.2% for us and has become genuinely stretched - the third straight session of momentum divergence at an RSI of 79, with the shares nearly 40% above their 200-day average. We are banking a third of the position into that strength and letting the rest run toward a still-healthy 12-week target. Note the 24-cent dip today is purely the 47-cent dividend going ex - on a like-for-like basis the shares actually rose, so this is a trim on extension, not on deterioration. Origination: ADR-0012 profit-take, armed Gain_ATR 4.27 (+10.2%, PROFIT_ARMED) + tick (a) third consecutive RSI_DIV at a RISING RSI 75.2->76.6->77.9->78.9, +39.7% over EMA200. G1 does not shield - AnticipationUp_VCP is not a BreakoutUp_*/MomentumContinuation_* family, so the softening clause never engages regardless of cv 78. Ticks (b)/(c)/(d) did not fire (conviction ROSE; T12 41.66 alive). Ex-div 47c captured today = Son. MoS -0.77 is not <= -3 so no tier bump, size_mod 1.0. 30% x 2,100 = 630 -> floor =. Realised +S (cost 28.63). f=0.138 on the 1wk. DCF annotation SUPPRESSED per SGO-3 (ResInc financial). Trail RAISED 30.41 -> 30.09 (a genuine +15c on the 29.94 ex-div-adjusted basis). Horizon: FOMC minutes 08-20, two-sided +-2.0% high-confidence - de-risking into a two-sided macro event is the conservative side of the catalyst-aware-hold clause, not a pre-trim into a favourable catalyst. Ladder -> -> 1,500 (26% of original still riding). · news
G07Great Eastern22.74SELL 30%22.7421.8923.21 (+2.1%)26.17 (+15.1%)31.99 (+40.7%)Great Eastern is our best position at +24.5% and also our most overheated - RSI 82, the highest in the book, with momentum now diverging from price, and today's 2.4% pop is a run into Wednesday's dividend that typically gives itself back on the ex-date. We are banking 30% into that run and keeping for what is still the strongest 12-week outlook we hold (+41%). Origination: ADR-0012 profit-take, armed Gain_ATR 8.34 (+24.5%, the most armed in the book) + tick (a) fresh RSI_DIV at RSI 81.9. G1 does NOT shield it - the signal IS MomentumContinuation_Up but conviction 71 is 9 points short of the 80 bar. IMPORTANT COLLISION: the same row prints BUY - ADD at adj 68 (Buy_Rank 1) - the add is SUPPRESSED by SGO-5 on both of its conditions at once (RSI 81.9 > 75, and the model entry 22.96 sits ABOVE the 22.74 close), and a name cannot be added to and trimmed in the same session; the mandatory ADR-0012 fire wins over the discretionary, gated-out add. Logged as systemic finding #2. MoS -0.68 not <= -3 so no tier bump. 30% x 1,200 = 360 -> floor =. Realised +S (cost 18.27), the book's best trim. Trimming at the 08-18 open forfeits 300 x 35c = Snear a wash against the mechanical ex-div mark-down. DCF blank -> annotation SUPPRESSED per SGO-3. Trail RAISED 21.41 -> 21.89, then steps to 21.54 on 08-19 (35c ex-div adjustment, NOT a lowering). Ladder ->. Never a full exit from profit-take alone. · news
D05DBS Group Holdings76.88HOLD77.2274.5978.12 (+1.2%)85.88 (+11.2%)95.75 (+24.0%)Holding all 7,700 DBS shares - our largest position at 29% of the book and up 2.9% - with the stop unchanged at S. Signal quality has faded three sessions running, so this is a watch, not a concern: the position is not extended enough to justify banking profit, and we are not adding into a 52-week-high breakout that this book's history says overshoots. Trail HOLDS at 74.59: today's printed stop 74.47 is LOWER and a trail never steps down. Conviction decayed -> 46 across three sessions and the row now carries RSI_DIV at RSI 72.0 - that 25-point slide WOULD be ADR-0012 tick (b), but D05 is NOT armed (Gain_ATR 1.58 < 3.0, T12_Progress 80.3% < 90), so it is a watch item, not a fire. Signal rotated to BreakoutUp_52W, which per L1 overshoots its 4wk targets - do not read the 85.88 as an add trigger. Q-fail non-binding per SGO-3; DCF annotation suppressed. Horizon: FOMC minutes 08-20 two-sided +-2.0% high-confidence, stance hold-through. f=0.138 on the 1wk.
S58SATS4.75HOLD4.754.614.78 (+0.6%)4.97 (+4.6%)5.69 (+19.8%)Holding all 60,200 SATS shares and nudging the stop up to S. The position is still slightly underwater at -2.1% but improving - price up 1.3% today, signal quality recovering, and the 12-week path intact - so nothing argues for cutting it. Trail RAISED 4.60 -> 4.61 (printed stop). Conviction recovered, T12_Progress 83.5% and rising, no Exit_Warning, not armed (Gain_ATR -0.92). Nothing originated a sell under any of the three §9 classes. f=0.138 on the 1wk.
S63ST Engineering11.13BUY-STOP11.009.98n/a13.05 (+18.6%)10.62 (-3.5%)ST Engineering is the closest thing to a buy in the book today and we are still standing aside. The signal is a promising early-stage setup, but the shares already trade far above what the fundamentals support, which knocks it just below our threshold - and our rules require the setup to appear twice before we will chase it, which it has not. On the watchlist for a second confirmation. Adj 64 = cv 74 + val_adj -10 (MoS -8.91), ONE point under the 65 floor. Separately INELIGIBLE for the ADR-0012 anticipate slot: BUY - ANTICIPATE printed in only 1 of the last 3 snapshots (needs >= 2), even though cv 74 clears the 55 slot floor. NOTE the 08-14 print is missing because the LIQUIDITY GATE MISFIRED on (systemic #1) - but a print that does not exist cannot be inferred, so the slot stays empty. Target_12Week 10.62 sits BELOW the 11.13 close, i.e. a dead forward view, which is an independent reason not to reach. Watch for a second ANTICIPATE print; size would be 0.5x normal new-BUY as a GTC buy-stop at the live model entry.

Outcome & track record

Accuracy at the time of this record.

25 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 →
0.138 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.

This website is for educational purposes only and does not constitute investment advice. Always do your own research and assess your own risk tolerance before making investment decisions.