AlphaBurst.ai
Independent research archive Hyperliquid
01 Research notebook / Volume 01

Research Notebook

Experiments in systematic finance.

An independent, evolving collection of computational-finance studies. Each experiment begins with a falsifiable question, documents its assumptions and data, and keeps simulated results beside the model's limitations.

Built with curiosity, not certainty. This is a developer's working laboratory for translating market ideas into code, tests, and reproducible evidence.

01
State the question

Turn an intuition about markets into a measurable, falsifiable hypothesis.

02
Test out of sample

Freeze assumptions, walk forward through time, and expose the model to unseen periods.

03
Document the failure modes

Keep bias checks, risk controls, and unresolved questions visible beside the result.

Walk-forward OOS Reproducible inputs Virtual capital Bias-aware

Just something for fun, built by an ex-Hedge Fund Quant Developer.

Connect with Me
Experiment index 1 documented study / 3 working papers
02 / Context Graph SynthesisFrozen MVP25 researchOpen research

Experiment 01 / Walk-forward study

Trend Continuation Under Guarded Risk

A directional options study that sells defined-risk spreads only when trend, strike location, liquidity, and portfolio controls agree.

Backtest period: Loading observed window...

Method: Train on three months, freeze parameters, then evaluate the next month out of sample using virtual capital.

$10,000 Study equity 1D -- | 7D -- | 30D --
Observed win rate--closed simulations
OOS max drawdown----
Observed return--walk-forward OOS
Information ratio--relative to SPY

Current experiment snapshot

Latest Option Portfolio as of --

Simulated positions; tile size reflects capital at risk.
Cached end-of-day marks
Loading current portfolio...

$10,000 strategy capital$5,000 weekly max-risk allocationAvailable risk = weekly limit - open risk - realized weekly losses

Open positions--
Capital at risk--
Max payoff--
Most likely payoff--Expected by POP
Open P&L--Current open positions
Realized P&L--Completed positions
Total Since Aug 14--Open + realized
Available risk--

Execution history

Option Trade Log

Waiting for the scheduled monitor.

Activity Ticker Entry IV Rationale Type of Trade Bull / Bear Entry Price Exit Price P&L Exit Reason
Loading trade history...

Scanner handoff

Bot Candidate Table

Loading latest bot candidates.

Ticker Sector Direction Spread Win Prob Credit Max Loss ROR Kelly Edge Kelly Cap Considered
Loading bot candidates...

Deferred risk exits

Stop Loss Warnings

No active stop-loss warnings.

TickerReasonCurrent MarkUnderlyingFirst WarningDeadlineEmailsStatus
No active stop-loss warnings.

Model diagnostics

Simulated Sector Risk vs SPY

Fresh-source snapshot loading...
SectorPositionsStudy riskSPY weightRelative
Building fresh risk attribution...

Sector exposure is normalized by gross open spread risk and compared with official SPY holding weights.

AlphaBurst.ai

Hyperliquid / Smart Trader Flow

Where smart money is moving

Collective position changes from the highest-ranked US stock-perp traders.

Loading flow
Flow window Loading coverage...
Collective inbound --
Collective outbound --
Net smart-money flow --
Observed coverage --

Buying / Long additions

Inbound

Cash entering
Loading inbound flow...

Selling / Short additions

Outbound

Cash leaving
Loading outbound flow...

Sector rotation

Sectoral shift

Tile size represents gross activity. Color and intensity represent net direction.

Net inbound Balanced Net outbound
Loading sector shifts...

Tracks public position changes from qualified wallets. Flow is observational, may be delayed, and is not investment advice.

A

AlphaBurst Research Notebook

Experiment 01: Trend Continuation

Hypothesis, walk-forward evidence, simulated positions, and risk diagnostics.

Updated --

Backtest Trading Days

Experiment Equity vs SPY

Simulated executions, open risk, realized P&L, and benchmark context.

$0 equity 0 names $0 risk $0 credit
Virtual equity SPY normalized Starting equity baseline Realized P&L exits V1 uses real underlying candles and mock option history until paid data is active.

Monte Carlo

$10k Safety Lab

Low sample
Median 10-90% range $10k
Median 90D $0
10th Percentile $0
Prob Below $10k 0%
Prob Below $7.5k 0%
Open P&L $0
Open Risk $0
Open Credit $0
Remaining Risk $0
Cumulative Realized P&L $0
Open Unrealized P&L $0
Cumulative Total P&L $0
Study Equity $0

Simulated Position Record

Selected Model Portfolio

ROR --

Ticker Status Direction Trade DTE Adj POP Kelly Edge Cap% Opened / Held Credit Risk P&L

Historic Portfolio

$0
Ticker Direction Trade DTE IV %ile Trend Jump POP Kelly Edge Cap% Opened / Held P&L

Strike Selection Analytics

How Did We Select Strikes?

Waiting for backtest rows.

$0 total

Experimental Observations

Simulated Trade Record