ψ Psi Quant Research

Signal Report · US

UPS HOLD

Composite decision report grounded in 1-XH strategy signals and a completed multi-agent debate (agents-3node-v1), generated at 2026-09-18 09:52.

Final Agent Decision

Action
HOLD
Weight
0%
Confidence
55%
Variant
xiaohan_gogo_v2

Reasons

Bull Case (Analyst)

- Technical signal is a clean multi-timeframe bottom: `gogo_detail` shows 1h/2h/3h/4h/daily all at bottom with `gogo_score` 0.70 and action=buy, and the cross-sectional `xh_gogo_score` z of +2.58 is by far the largest lifter — the same signal that drives the trade is the dominant factor in the rank.
- Composite z = +0.29, rank 63/252 (top 25%), in a BULL catalog with 0 regime direction-flips, so the factor tilt is aligned with the long side even if not top-decile.
- Volatility factors corroborate a bottoming setup rather than a breakdown: `downside_vol_60d` (+0.34) and `volatility_60d` (+0.11) are positive contributors, consistent with a washed-out name where realized vol is being rewarded.
- News flow is not uniformly negative: 1-year total shareholder return +25.59%, a +1.42% outperformance day vs the S&P 500 (MarketWatch), and a $2B investment program in international/healthcare/supply chain (CNBC) — plus a 29,223 RSU grant to the chief digital officer, a routine insider alignment item.

Bear Case (Analyst)

- The cross-section explicitly did not select it: UPS is **not in the production basket**, and rank 63/252 is top-quartile, not top-decile — the composite is only mildly positive, so this is a marginal long at best.
- The draggers are all trend/positioning factors that argue the bounce is counter-trend: `close_to_high_60d` (-1.24), `bollinger_b_20d` (-0.72), `sharpe_60d` (-0.64) — price is near the low end of its 60-day range with weak risk-adjusted returns.
- Price is below both the fast and slow EMAs (close 99.01 vs ema_low_fast 99.62 and ema_high_slow 101.50), and CNN notes UPS trades below its 200-day SMA and mid 52-week range — the "bottom" signal is fighting a downtrend, not riding one.
- Momentum is negative on the intermediate horizon: 30-day return -3.69% and 90-day -5.53% (Simply Wall St), with the stock -1.11% on the signal date.
- Fundamental overhang from Motley Fool: fuel surcharge change swung from -$280M (2024) to +$1,173M in H1 2026, a large earnings-quality/expense dynamic that the technical bottom signal does not price.

Note: the technical signal (buy) and the factor rank (top 25%, not basket) partially contradict — the basket exclusion is the strongest single piece of evidence against sizing this as a conviction long.

Devil's Advocate

COUNTER TO BULL:
- The "clean multi-timeframe bottom" is one signal counted five times. 1h/2h/3h/4h/daily all bottoming is not five confirmations, it's one oscillator at five resolutions — correlated, not independent. The bull treats redundancy as confluence.
- `xh_gogo_score` z of +2.58 being "the dominant factor in the rank" is a red flag, not a feature. If one cross-sectional factor drives the rank, the composite z of +0.29 is largely that single factor wearing a costume. Strip it out and the rank likely collapses below top-quartile.
- "BULL catalog with 0 regime flips" says nothing about this name. A regime label is a market-wide backdrop; it does not validate a marginal, non-basket long. Alignment with the long side is the null hypothesis, not evidence.
- Positive `downside_vol_60d` and `volatility_60d` contributions are being read as "washed-out bottom." They are equally consistent with "high realized vol, unstable, uninvestable." The bull asserts the interpretation it needs.
- The news items are non-causal. +25.59% 1-yr TSR is backward-looking and describes the top you're now buying below. A +1.42% outperformance day is noise. A $2B investment program is a cost, not a catalyst, until it shows up in margins. An RSU grant is dilution/comp, not insider conviction — no open-market buy.
- Hidden assumption: that a bottom signal on a downtrending name is a reversal rather than a pause. No evidence of a higher low, no volume confirmation, no trend break cited.

COUNTER TO BEAR:
- "Not in the production basket" is a statement about the model's selection, not about the trade's edge. Basket membership is a threshold artifact; rank 63/252 is still top-quartile and the bear offers no evidence the cutoff is where alpha dies.
- The draggers (`close_to_high_60d`, `bollinger_b_20d`, `sharpe_60d`) are all mean-reversion-friendly when negative. Calling them "counter-trend" assumes trend persistence — the exact assumption a bottom signal is designed to fade. The bear is using trend logic to dismiss a mean-reversion signal; that's a category error.
- Below EMAs and below 200-day SMA is the definition of a bottoming setup, not a refutation of one. If price were above the EMAs, the "bottom" claim would be false. The bear's strongest-sounding point is circular.
- Negative 30d/90d momentum is the precondition for the signal, not a contradiction of it. Citing it as a counter-argument double-counts the same information the bull already priced.
- The fuel surcharge swing is a real fundamental item but the bear doesn't quantify it against earnings, guidance, or consensus. A -$280M→+$1,173M swing is actually a *tailwind* to reported earnings in H1 2026 unless the bear shows it's non-recurring or offset. As stated, it may support the bull.
- Hidden assumption: that "marginal long" is not worth taking. No position-sizing or expected-value argument is made — the bear conflates low conviction with negative edge.

Technical Signal (1-XH)

action
buy
score
0.70
variant
xiaohan_gogo_v2
timestamp
2026-09-18T15:30:00-04:00
close
99.01
ema_low_fast
99.62
ema_high_slow
101.5
gogo_score
0.7
gogo_detail
1h_bottom,2h_bottom,3h_bottom,4h_bottom,daily_bottom

📈 Backtest Track Record xiaohan_gogo_v2 · signals simulated over available history (~3-month 30-min window, accumulating daily)

Trades
4
Win rate
50%
Avg / trade
+2.2%
Total (compounded)
+9%
best / worst
+10.7% / -1.5%
avg holding
16.2 days
open position
yes — marked-to-market -1.5%
Entry Exit Entry Exit Return Exit reason
2026-09-17 open 100.55 99.01 -1.5% open_position_mark_to_market
2026-09-01 2026-09-01 102.85 102.46 -0.4% adaptive_breakdown_1h
2026-08-19 2026-09-01 103.41 103.47 +0.1% adaptive_breakdown_30m
2026-05-19 2026-07-09 100.16 110.83 +10.7% daily_sell_after_daily_blue_cross

复盘 = each past xiaohan_gogo_v2 buy signal simulated: enter next bar, manage adds / adaptive sells / −25% hard stop. Demonstration only, not investment advice.

News & social (8 items)