Historical research only. Not investment advice, not a promise of future returns, and not a broker order.
ClosedUpdated 4:27:15 PM Eastern
How’d the model do today?
The model is −0.05% today, −0.79% versus the available passive benchmark.
Paper Model Profit / Loss
−$53.82
−0.05% Today
Model Today
−0.05%
−$53.82
SPY Today
+0.74%
$741.42
VTI Today
+0.74%
Benchmark Proxy
Edge Vs VTI
−0.79%
Model Minus VTI
ModelSPYVTI10:10 AM–3:58 PM Eastern
Latest Model Actions
A compact view of the latest paper fills that contributed to today’s result.
LRCXSellAdaptive Signal$9,933.35
LRCXBuyAdaptive Signal$9,987.17
Recent research results
What a $100K research account would have done
Class
Public Model
Dates
Observations
Starting Capital
Modeled P/L
Return
Ending Value
Evidence
Neural Router
Router Cohort 1
May 6, 2026 – Today
61,678
$100,000
▲+$384,812.03
+384.81%
$484,812.03
Trade Ledger
Combo + Meta
Router Rule Set 4
May 6, 2026 – Today
38,112
$100,000
▲+$94,210.88
+94.21%
$194,210.88
Validated
Combo Algo
Strategy Pair 1
May 6, 2026 – Today
22,904
$100,000
▲+$72,440.00
+72.44%
$172,440.00
Simulation
Mono Algo
Strategy 1
May 6, 2026 – Today
10,742,752
$100,000
▲+$25,515.83
+25.52%
$125,515.83
Full History
Read it like this: if the modeled $100,000 research account followed the listed model over the listed date range, this is the ending value the simulation produced.
These are historical simulations, not live broker returns, investment advice, or future-return promises.
Research Preview
A public glimpse of the lab without exposing the edge.
Visitors can see the structure: leaderboards, model classes, benchmark comparison, and proof-ledger workflow. The proprietary names, scores, and exact statistics stay locked until they have access.
Subscribe to reveal exact strategy names, model scores, and exportable trade ledgers.
Historical Leaderboard
All completed research windows
ModelsLedgers
Historical Leader
███ ██████
+██.██%
Neural Approval
Model █
██.█%
SPY Comparison
Today
±█.██%
VTI Comparison
Today
±█.██%
Rank
Class
Model
Dates
Return
1
Neural
Cohort █
May 6 – Today
+███.██%
2
Combo + Meta
Rule Set █
May 6 – Today
+██.██%
3
Mono Algo
Strategy █
Full Window
+██.██%
4
Validation
Context █
Holdout
+█.██%
How Leroy Works
A shorter path from idea to evidence.
Compare the result, verify the path, then decide whether the signal deserves more attention. The page keeps the details available without turning the pitch into documentation.
Single signals, combinations, meta rules, and adaptive models compared side by side.
External Sensors
Expanding
Kalshi-style event markets, macro, sector, breadth, volatility, and earnings context.
Trade Ledgers
Line By Line
Every headline result can be checked against the modeled buys and sells behind it.
Today
Know what deserves attention before the next session.
01 / 05
The product answer is practical: it turns a pile of strategies, simulations, and model scores into a ranked short list you can inspect today.
Question
What Worked?
Proof
Ledger + Path
Next Move
Paper Or Review
Start with the date range you care about.
See which class led and why it qualified.
Use the evidence to choose what to test next.
Compare
One leaderboard for every strategy class.
02 / 05
Start with the practical question: which approach held up over this date range? Single algos, combinations, meta rules, and neural cohorts only compete when the coverage is comparable.
View
Leaderboard
Classes
Mono · Combo · Meta · Neural
Evidence
Complete Windows Only
Rank results across the same dates.
Hide incomplete comparisons.
Keep proprietary strategy names private.
Verify
Every return can be checked trade by trade.
03 / 05
A big number is not enough. Leroy keeps the result tied to the modeled buys, sells, gains, losses, and balance path that created it.
Start
$100,000
Check
Trade Ledger
Output
Line-By-Line Proof
Open the ledger behind the return.
See the running balance change.
Separate evidence from screenshots.
Adapt
Strategies become sensors. The router chooses the context.
04 / 05
The goal is not to worship one strategy forever. The stronger system asks which signal mix historically worked best under similar market conditions.
Signals
Breakout · Reversion · Momentum
Context
Time · Trend · Volatility
Model
Meta + Neural
Score single and combined signals.
Learn from rejected setups.
Promote only after validation.
Expand
More independent sensors, not more noise.
05 / 05
The roadmap layers in external context like Kalshi-style event markets, macro calendars, sector pressure, breadth, volatility, and earnings risk.
Live
Supervised Workflow
Sensors
Market + Event Context
Research
Faster Reruns
Add non-price inputs.
Backtest new layers before use.
Keep paper research running beside live mode.
Research Access
Start with the leaderboard. Audit with the ledger.
Claim $100 lifetime founding access while spots are available, or subscribe to realtime data access for $20/month.