ψ Psi Quant Research

Signal Report · US

OKLO HOLD

Composite decision report grounded in 1-XH strategy signals and multi-agent debate, generated by DeepSeek at 2026-08-11 09:06.

🎯 Final Decision

Action
HOLD
Weight
0%
Confidence
40%
Variant
xiaohan_gogo_v2

Reasons

Bull Case (Analyst)

- Technical signal is a clean buy (score 0.70) with the 1-XH family showing bottoming across 1h–4h timeframes and EMA fast (44.42) above slow (43.94) at close 44.51 — momentum is turning up from a local base.
- Cross-sectional factor view supports a tactical long: xh_gogo_score z = +2.68 is the top lifter, and beta_60d z = +1.71 adds high-beta exposure — in a BULL regime catalog with zero direction-flips, this combination favors upside participation.
- News flow is strongly positive: Trump administration selected Oklo for a $200M federal initiative to power AI data centers (barchart, stocktwits), and the Groves Isotope Test Reactor achieved first criticality on private land — a DOE pilot milestone (stocktitan).
- Analyst price target of $79.56 implies ~89% upside from current levels (tradingview), and the stock rallied 14% to $48.21 on first meaningful quarterly revenue (247wallst) — fundamental catalysts are stacking.
- Insider selling of $4.9M by a 10% shareholder (moomoo) is small relative to the $200M federal push and reactor milestone — likely noise, not a signal of deteriorating fundamentals.

Bear Case (Analyst)

- Cross-sectional rank is mediocre: composite z = +0.01, rank 132/252 (52.4%) — the stock is not in the production basket and the factor model sees no edge here; the technical buy is not corroborated by cross-sectional strength.
- Top draggers are severe: drawdown_3m z = -2.00, momentum_6m z = -1.45, sharpe_60d z = -1.43 — the stock has been a persistent underperformer with poor risk-adjusted returns, and the BULL regime catalog does not flip these negatives.
- Price action is fragile: stock is down 28% in a month (yahoo), down 33% YTD (247wallst), and down 41% YTD per stocktwits — the technical buy may be a dead-cat bounce within a longer downtrend.
- Insider selling is a red flag: a 10% shareholder sold 120K shares for $4.9M on Aug 5 (moomoo) — insiders are monetizing the recent rally, which often precedes further downside.
- The 12-month price target was cut to $79.56 (tradingview) — even the bullish case is being revised downward, and the stock's high beta (z = +1.71) cuts both ways: in a risk-off tape, OKLO will fall harder than the market.

Devil's Advocate

COUNTER TO BULL:
- The "clean buy" is built on a 1-XH family that has been wrong repeatedly in this name — the same bottoming signal fired during the -28% month. EMA spread of 0.48 on a $44 stock is statistically meaningless; it's noise around a downtrend, not a reversal.
- The factor view is cherry-picked: xh_gogo z=+2.68 is a short-horizon momentum oscillator that mean-reverts violently. You're pairing it with beta_60d z=+1.71, which is a *risk* loading, not an alpha signal — in a BULL regime catalog that hasn't been validated out-of-sample for this ticker.
- News catalysts are non-revenue: $200M federal initiative is a *potential* award, not a contract; first criticality is a technical milestone with zero commercial revenue attached. The 14% pop on "first meaningful quarterly revenue" was from a base of essentially zero — absolute numbers are still tiny.
- The $79.56 target is a consensus lag — it was likely set before the recent collapse and is being cut (see bear case). Anchoring on a stale target while the stock trades 44% below its high is naive.
- Insider selling of $4.9M by a 10% holder is *not* noise at this market cap — it's ~1% of the float and was executed into strength. Insiders sell for many reasons, but buying the dip after a 14% rally on news is a bet against their information.

COUNTER TO BEAR:
- Composite z=+0.01 and rank 132/252 is *not* mediocre — it's dead center, meaning the factor model is indifferent, not negative. Using "not in the production basket" as bearish is circular: the basket is a trading rule, not a fundamental verdict.
- The draggers (drawdown_3m, momentum_6m, sharpe_60d) are all *backward-looking* — they describe the pain already realized. If the stock has bottomed (as the technicals suggest), these will mechanically improve; using them to argue against a reversal is recency bias.
- The -28% month and -33% YTD are facts, but they're also the setup for a mean-reversion long. Calling the technical buy a "dead-cat bounce" requires evidence the downtrend is intact — the 1h–4h bottoming and EMA cross are at least *some* evidence it isn't.
- Insider selling: one 10% holder selling $4.9M is a data point, but it's not a controlling shareholder, and the sale was likely for liquidity/estate reasons. Without a pattern of repeated insider distribution, this is weak sauce.
- The price target cut to $79.56 still implies 89% upside — a cut from a higher fantasy number to a still-fantasy number is not a bearish signal. High beta cuts both ways: in a risk-on tape (which the BULL regime catalog suggests), OKLO will outperform.

Technical Signal (1-XH)

action
buy
score
0.70
variant
xiaohan_gogo_v2
timestamp
2026-08-10 15:30:00
close
44.51
ema_low_fast
44.42
ema_high_slow
43.94
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
50%
Avg / trade
-9.0%
Total (compounded)
-20%
best / worst
+7.4% / -25.4%
avg holding
30.0 days
open position
yes — marked-to-market +7.4%
Entry Exit Entry Exit Return Exit reason
2026-07-16 open 41.45 44.51 +7.4% open_position_mark_to_market
2026-06-11 2026-07-16 56.38 42.05 -25.4% hard_stop_-25%

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)