ψ Psi Quant Research

Signal Report · US

CSGP HOLD

Composite decision report grounded in 1-XH strategy signals and multi-agent debate, generated by DeepSeek at 2026-07-20 08:53.

🎯 Final Decision

Action
HOLD
Weight
0%
Confidence
30%
Variant
xiaohan_gogo_v2

Reasons

Bull Case (Analyst)

- Technical signal is outright bullish: the 1-XH family flags a "buy" with a score of 0.70, and the gogo_detail shows 1h/2h/3h/4h bottoms plus a blue-upper-above-yellow-upper configuration — a multi-timeframe reversal pattern.
- The cross-sectional factor view ranks CSGP in the top 15.1% of the universe (rank 38/251) with a composite z of +0.30, placing it in the upper tier of long candidates in today's BULL regime catalog.
- The strongest factor lifters are xh_gogo_score (+2.94 z) and bollinger_b_20d (+0.56 z), directly corroborating the technical buy signal — the same gogo model that triggered the buy is the #1 factor contributor to the cross-sectional long case.
- Mean_revert_20d (+0.40 z) suggests the stock is oversold on a 20-day basis, aligning with the 55.4% YTD decline cited in news — mean reversion potential from deeply depressed levels.
- News of France expansion and steady U.S. office leasing (Q2 data from CoStar) provide fundamental catalysts that could support a turnaround narrative, despite the year-to-date selloff.

Bear Case (Analyst)

- The cross-sectional factor view reveals severe drag from momentum and risk-adjusted return factors: sharpe_60d (-1.58 z), close_to_high_60d (-0.85 z), and momentum_6m (-0.76 z) are all strongly negative — the stock has poor recent risk-adjusted performance and is far from its highs.
- CSGP is NOT in the production basket — the cross-section explicitly excluded it from the top decile this period, meaning the systematic quant model does not see it as a top long despite the composite z being positive.
- News sentiment is overwhelmingly negative: shares have plunged 55.4% YTD, retail rent growth slowed to 1.6% (slowest in a decade per Q2 data), and the only legal news involves a disqualification motion — no positive earnings or revenue catalysts.
- The technical buy signal (score 0.70) is contradicted by the factor view: the gogo score is the sole strong lifter, but it's being overwhelmed by three momentum/drawdown factors that are deeply negative — the composite z of +0.30 is weak and barely positive.
- The France expansion news is a long-term strategic move, but with the stock down 55% YTD and rent growth decelerating, near-term headwinds from the core U.S. commercial real estate market likely outweigh any international expansion narrative.

Devil's Advocate

COUNTER TO BULL:
- The gogo score (z +2.94) is the *only* strong lifter; the composite z of +0.30 is barely above zero, meaning the rest of the factor stack is neutral-to-negative. A single technical model does not a long make.
- The 1-XH buy signal (0.70) is not a high-conviction threshold — it's barely above the typical 0.5 cutoff. Multi-timeframe bottoms can be false signals in a persistent downtrend, especially when momentum factors are deeply negative.
- Mean_revert_20d (+0.40 z) is weak and could simply reflect a dead-cat bounce in a stock that has further to fall — oversold does not mean cheap, it means the trend is your enemy.
- France expansion is a multi-year narrative with no near-term revenue impact; citing it as a catalyst for a 55% YTD loser is narrative-driven, not data-driven.
- The stock is not in the production basket — the systematic model explicitly passed on it. If the quant desk's own production process doesn't back it, why should the analyst?

COUNTER TO BEAR:
- The composite z of +0.30 is positive, and rank 38/251 is top 15% — that's not "barely positive," it's solidly in the upper tier of the BULL regime catalog. The model *does* see it as a candidate, just not top-decile.
- The gogo score (+2.94 z) is the strongest factor lifter across the entire cross-section for this stock — dismissing it as "the sole strong lifter" ignores that it's an outlier signal with high statistical weight.
- Momentum factors (sharpe_60d, close_to_high_60d, momentum_6m) are backward-looking; the technical buy signal is forward-looking. A reversal pattern by definition occurs *after* momentum has deteriorated — that's the point.
- The 55.4% YTD decline is already priced in; mean_revert_20d (+0.40 z) suggests the selling has exhausted on a short-term basis. News sentiment is stale — the market looks forward, not backward.
- Not being in the production basket is a threshold decision, not a rejection. The composite z is positive; the model just didn't rank it high enough this period — that's a marginal call, not a bearish signal.

Technical Signal (1-XH)

action
buy
score
0.70
variant
xiaohan_gogo_v2
timestamp
2026-07-17 15:30:00
close
29.78
ema_low_fast
29.26
ema_high_slow
29.57
gogo_score
0.7
gogo_detail
1h_bottom,2h_bottom,3h_bottom,4h_bottom,1h_blue_upper_above_yellow_upper

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

Trades
2
Win rate
0%
Avg / trade
-13.3%
Total (compounded)
-25%
best / worst
-12.8% / -13.8%
avg holding
54.6 days
open position
yes — marked-to-market -12.8%
Entry Exit Entry Exit Return Exit reason
2026-05-05 open 34.14 29.78 -12.8% open_position_mark_to_market
2026-03-17 2026-04-22 44.51 38.37 -13.8% adaptive_breakdown_1h

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 (7 items)