SG Off-the-Radar — End-of-Day Verdict
0 ACQUIRE + 2 DISPOSE. Exactly one row cleared the actionable bar - O39 BUY - ADD cv73 at Buy_Rank 1, adj 69 - and it was SUPPRESSED per SGO-5 on both conditions at once (RSI 76.62 > 75 AND Entry_Price 31.67 >= Close 31.50) plus the no-flip-flop rule, the book having sold of O39 at this morning's open. DISPOSE: O39 OCBC TRIM @ S - ADR-0012 profit-take, armed at Gain_ATR 3.96 (+10.02%) with tick (a), a fresh RSI_DIV at RSI 76.6 on a new 52-week high 38% above EMA200; G1 misses by seven conviction points (73 vs the 80 bar). Banks S / Srealised and keeps behind a trail raised to 30.22 - the second rung of a de-risking ladder -> 2,900) that leaves half the original position riding a still-positive 30% 12wk path. J36 JARDINE MATHESON TRIM @ S - the first cut, on §9 'N consecutive NoTrade/Conflicted' at N=4 with the 12wk falling every session to -11.6%; the PROFIT_ARMED is a false positive on a losing position and the escalation clause did NOT fire (close 61.93 vs 61.76, though the day's low pierced it for the first time), so the full-exit clause stays live on the remaining, ex-div-adjusted to 61.11 on 08-20. UNDERVALUED MoS +0.38 holds the tier at the 30% floor; books a Sloss. G07 is the most armed position in the book (Gain_ATR 6.92) but NOT fired - no tick qualifies, so trail-raise 21.11 -> 21.33 only. D05 holds 7,700 into tomorrow's 81c ex-div with the stop lifted to a dividend-adjusted 74.59; S58's headroom narrowed to 1.7%, the tightest in the book. Four of five stops raised; zero trail erosion. Trims free Staking to S and to 52.7% - materially under-, awaiting a genuine PullbackBuy setup rather than a lower bar. Systemic: the engine emitted BUY - ADD and PROFIT_ARMED on the SAME O39 row - it recommended adding to a position its own exit layer had armed.
Recommendations
0 to acquire · 2 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 |
|---|---|---|---|---|---|---|---|---|---|
| O39 | OCBC Bank | 31.50 | SELL 30% | 31.50 | 30.22 | 32.12 (+2.0%) | 36.01 (+14.3%) | 40.97 (+30.1%) | OCBC has run 10% above our cost to a fresh 52-week high and is now 38% above its 200-day trend line with RSI at 77, and a bearish momentum divergence has just appeared alongside the profit-take arm. Record 1H results (net profit +13% to S19bn, interim dividend raised 15% to 47 cents) are the reason the stock is here - they are already in the price, not ahead of it - so we bank a third of the remaining stake into the strength and keep for the still-positive 30% twelve-week path. ADR-0012 profit-take: armed (Gain_ATR 3.96 on a real +10.02% gain) + tick (a), a fresh RSI_DIV code on the same row. G1 does not shield it - the family qualifies (fresh MomentumContinuation_Up) but conviction is 73, seven points short of the 80 bar; at cv80 this would have been a trail-raise instead. Base 30% tier (1 tick; MoS -0.77 is not <= -3 so no valuation bump). 1,230 rounded DOWN to the 100-share board lot. The SAME row prints BUY - ADD at Buy_Rank 1: SUPPRESSED per SGO-5 on both conditions at once (RSI 76.62 > 75 AND Entry_Price 31.67 >= Close 31.50), and by the ACT-NOW no-flip-flop rule - the book sold at this morning's open, so adding at 31.67 would be paying up 1.8% to repurchase it. Action price 31.50 is the settled close, the floor of the provenance band [31.49, 31.85] (Entry 31.67 +- 0.25 x ATR 0.7241); the printed 31.67 is a buy-stop above the close and is not a level a seller can rely on. Retained, trail RAISED 29.73 -> 30.22 (printed model stop). Ex-div 47c on 08-17: selling on the 08-14 open is pre-ex, so it is mechanically a wash - forfeits Scaptures the pre-div price; retained shares collect SSGO-3 suppresses the DCF-anchor flag and the Quality_Pass=False block (IsFinancial, FV=ResInc). Re-arm for a buy: a BUY/STRONG BUY print with RSI back below 75 and the model entry BELOW the close. · news |
| J36 | Jardine Matheson | 61.93 | SELL 30% | 61.93 | 61.76 | 61.70 (-0.4%) | 60.23 (-2.7%) | 54.78 (-11.5%) | Jardine Matheson has now gone four straight sessions with the model authorising nothing constructive - two blanks then two outright bearish prints with conviction rising - while the position sits at a loss, below both its 10-day and 200-day trend lines, and the twelve-week view has fallen every single session to 11.6% BELOW today's price. The business itself is fine (1H underlying profit +9%, free cash flow +21%, dividend raised 8%), and it is the only cheap name we own, which is exactly why we cut the minimum 30% rather than exiting: we stop a deteriorating chart from compounding without dumping a sound, undervalued holding. Origination is CLAUDE.md §9 'N consecutive NoTrade/Conflicted on a held position' at N=4 (NO TRADE 08-10, NO TRADE 08-11, AVOID 08-12, AVOID 08-13), corroborated by Target_12Week falling monotonically 58.36 -> 57.36 -> 57.12 -> 54.78. NOT originated by ADR-0012: the PROFIT_ARMED is a FALSE POSITIVE for the fourth session - the position is at a LOSS (Gain_ATR -0.95) and the arm comes solely from a degenerate T12_Progress 113.05 that reads >100% only because Target_12Week 54.78 sits BELOW the price. NOT the escalation clause either: the settled close 61.93 is 17 cents ABOVE the 61.76 line, though the day's LOW 61.72 pierced it for the first time (08-12's low missed by two cents) - L4 makes the settled close authoritative, so this is a partial trim and the full-exit clause stays LIVE on the remaining. Valuation amplifies DOWN one tier (UNDERVALUED, MoS +0.38, Quality_Pass True) - 'selling cheap, trim the minimum' - holding the tier at the 30% floor. Acted now rather than deferring a fourth time per L3 and the [saved] optional_trim counterfactual (median 1wk +1.12%, n=6); the book deferred on 08-06/08-11/08-12 on a few cents each time while the 12wk fell 3.6 points. 510 rounded DOWN to board lot. Books a Srealised loss. Action price 61.93 = the printed Entry_Price exactly, band [61.50, 62.36]. The row's printed Stop_Price 65.34 is the SHORT-side stop of RallySell_BearTrend and is NOT applicable to a long. MANDATORY EX-DIV STOP ADJUSTMENT: J36 goes ex 65 US cents on 08-20 - on that date the stop steps mechanically to 61.11 (61.76 - 0.65). That is a dividend adjustment, not a lowering; without it the escalation clause would fire spuriously on the mark-down alone. · news |
| G07 | Great Eastern | 22.27 | HOLD | — | 21.33 | n/a | n/a | 31.58 | Great Eastern is the book's biggest winner at +21.9% and its most extended name at RSI 80, but nothing deteriorated today - so we tighten the stop rather than sell more, having already banked at this morning's open. Strong 1H results (net income S, 35-cent interim dividend) support riding the remaining behind a closer line. ARMED (Gain_ATR 6.92, the most armed position in the book) but NOT FIRED - no tick qualifies: (a) the row carries PROFIT_ARMED alone, the 08-11 RSI_DIV has not returned; (b) yesterday's cv32 -> today's cv0 is NOT a 20-pt conviction collapse - a NO TRADE row prints no conviction, entry, stop or target at all, and reading its cv0 as a graded score is the artifact SGO-4 warns against; (c) Target_12Week 31.58 vs 1.02 x 22.27 = 22.72 is far from dead; (d) this is the FIRST NO TRADE of the run, not the second (08-12 printed HOLD). Armed-but-not-fired is a MANDATORY trail-raise, not a trim. New trail 21.33 = chandelier at 2.0 x ATR (0.5778) below the session high 22.49; a raise from 21.11, so ADR-0020 peak = live, 0% erosion. Ex-div 35c on 08-19 - per the 08-11 ruling the mark-down must NOT be read as deterioration next week. · news |
| D05 | DBS Group Holdings | 76.50 | HOLD | 76.50 | 74.59 | n/a | 79.30 (+3.7%) | 95.18 (+24.4%) | DBS carries the strongest twelve-week outlook in the book (+24% to S), conviction has recovered from, and it pays an 81-cent dividend tomorrow - so we hold all and simply lift the stop. NOT ARMED on either profit-take path (Gain_ATR 1.28 on the true blended cost S, T12_Progress 80.37 - needs >= 3.0 or >= 90), Exit_Warning blank, so no trim is available. Today's printed model stop 75.40 is a large raise from 73.54, but at 0.79 x ATR below the close it is far tighter than a 2-ATR trail AND D05 goes EX-DIVIDEND 81 cents tomorrow, 08-14, the execution date (66c ordinary + 15c capital return). Adopting 75.40 unadjusted would leave the stop within 0.29 of the post-mark-down price and risk a purely mechanical stop-out on the book's best forward view. Stop therefore set to 74.59 = 75.40 - 0.81, the dividend-adjusted equivalent of the model's own line - still a 1.05-point RAISE over the recorded 73.54 and immune to tomorrow's mark-down. Holding captures 7,700 x S = SSGO-3 suppresses the DCF-anchor flag and the Quality_Pass=False block (IsFinancial, FV=ResInc). · news |
| S58 | SATS | 4.68 | HOLD | 4.74 | 4.60 | n/a | 5.09 (+7.4%) | 5.71 (+20.5%) | SATS is down 3.5% and its price has drifted to within 1.7% of our stop - the tightest margin in the book and the position most likely to force a decision next session - but nothing has originated a sell: the signal is a stable four-session HOLD, not a REDUCE, and the twelve-week view still points 22% higher. We hold and lift the stop to the model's line. NOT ARMED (Gain_ATR -1.57), no Exit_Warning. Conviction slipped, a 9-point move well short of the 20-point deterioration tick. Trail RAISED 4.59 -> 4.60 (today's printed model stop); ADR-0020 peak = live, 0% erosion. Headroom narrowed from 2.5% to 1.7% in one session - watch closely. T12_Progress is actually RISING (80.5 -> 82.0). FAIR_VALUE at MoS -2.42 is the miscalibrated banding note B warns about - read cautiously; it originates nothing (§9). · news |
| S68 | Singapore Exchange | 24.97 | WATCH | 24.97 | 24.81 | n/a | 26.01 (+4.2%) | 27.74 (+11.1%) | SGX is the strongest unheld name on the board and the only one still improving, but the model has yet to print an actual buy trigger on it - a WATCH is a monitoring state, not a signal - and on our valuation work it is expensive here. We stay flat and wait for the trigger. WATCH cv63 (up from 60) is not an actionable Action, so the floor's first clause fails before conviction is even scored; val_adj -9 on OVERVALUED_VS_RI MoS -1.84 takes it to adj 54 regardless. Fifth consecutive session without a printed BUY. Re-arm trigger: a printed BUY/STRONG BUY - then queue at the model Entry_Price per SGO-5, never a deeper self-invented limit. |
| C07 | Jardine C&C | 27.29 | WATCH | — | — | n/a | n/a | 24.40 | Jardine C&C remains the cheapest quality name on our screen, but cheapness alone has never been a reason to buy in this book - and today the technical case walked away entirely rather than toward us. We stay out. The signal collapsed WATCH cv51 -> NO TRADE cv0 ('Consensus conflict + low conviction'), the close fell 2.1% to 27.29, RSI dropped to 43.9, and the price now sits below its 10-day (27.98) and far below its 200-day (29.72) trend line. Its +10 val_adj (UNDERVALUED, MoS +0.61, Quality_Pass True) is the best fused valuation score in the book and still leaves adj 10 with no signal to attach it to. OTR-2 is the governing lesson: UNDERVALUED without a signal is not a buy. Wait for the technical trigger. |
| 9CI | CapitaLand Investment | 2.75 | WATCH | 2.65 | 2.57 | n/a | 2.84 (+7.2%) | 2.94 (+10.9%) | CapitaLand Investment has already run away from the price the model wants to pay, and its conviction is fading session by session - buying here would be chasing. We stay out and wait for the pullback. WATCH cv50 (fading -> 50), val_adj -9 on GROWTH_PRICED_IN MoS -1.87, adj 41 - well under the 65 floor. Entry_Distance_ATR -1.70 puts the close 1.7 ATR PAST the model entry (2.65) with RSI 69.6: CLAUDE.md §9's wait-for-pullback shape. Re-arm: a BUY print with the close back at or below the model entry. |
Outcome & track record
Accuracy at the time of this record.
21 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.