Hypotheses

Derived by the Research Lead from Ethereum staking-withdrawal mechanics: test whether an unusually large wave of newly initiated voluntary full exits predicts ETH underperformance over the following five days. Ethereum documents that full exits unlock validator balances for execution-layer use; it does not claim that withdrawals cause sales or prescribe this strategy ([Ethereum staking withdrawals](https://ethereum.org/staking/withdrawals/)). Trade ETHUSD.HYPERLIQUID short and SOLUSDT.BYBIT long using synchronized 4-HOUR bars. This is a supply-intention event hypothesis, not a price-shock fade, funding-carry trade, or continuously rebalanced ratio strategy. Orders execute at completed signal-bar closes, the registry equivalent of next-bar-open execution on continuous markets. Protection is a platform-matched intrabar stop. Historical chain data is required before this hypothesis can be tested; no substitute based on price, funding, or the change in queue length is allowed. No preliminary backtest or event-count measurement has been performed.

Derived by the Research Lead from Interview with Atsushi Yamanaka (ATC 2011), https://www.mql5.com/en/articles/544: test the disclosed idea of concentrating capital in specialists ranked by historical profit factor and win rate, with Parabolic SAR trailing exits. The complete locally archived interview was read. It does not disclose reproducible entries or ranking windows; the breakout experts, sparse eligibility gate, cost treatment, and allocation rules below are Research Lead choices, not Yamanaka's strategy. Use completed UTC 4-HOUR bars. Execute ordinary market orders at the signal bar close on this 24/7 venue, the registry equivalent of next-bar-open execution. Maintain separate causal shadow histories for every specialist, including unallocated ones, so allocation does not censor future eligibility. This is a research prediction, not a claim of a demonstrated edge. The central test is whether specialist selection improves on identical experts with constant equal allocation.

Original Research Lead hypothesis. Trade short-only event positions in AAVE, UNI, LINK and AVAX Hyperliquid perpetuals using synchronized 4-HOUR bars and matching Binance spot bars as signal sources. Test whether a large upside shock followed by an unsuccessful retest and a break below the shock midpoint identifies trapped buyers and subsequent inventory liquidation. This is a delayed failure trade, not an immediate squeeze fade. Orders execute at the signal bar close on this continuous venue; this is the registry's faithful equivalent of next-bar-open execution. Use platform-matched intrabar protective stops. No exchange-hosted protection, queue advantage or selectable same-bar policy is required.

Original Research Lead hypothesis: test whether an abrupt increase in ETH options-implied uncertainty, accompanied by a material same-direction move in ETH spot, predicts directional continuation over the following 72 hours. Trade ETHUSD_PERP.BINANCE_CM long or short; ETHUSDT.BINANCE_SPOT is an untraded confirmation source. Use UTC-aligned 4-HOUR bars and the point-in-time supplementary DVOL_ETH series. This trades directional information after an options repricing event, not options, variance-premium levels, an ETH/BTC ratio, or volatility-state stop-and-reverse. Continuous-market orders execute at the signal bar close, the registry-approved equivalent of next-bar-open execution. Inverse sizing and platform-matched intrabar protective stops are supported. No exchange-hosted protection, hedge mode, selectable same-bar policy or latency advantage is required. All return and event-rate figures below are preregistered predictions, not measurements; no preliminary backtest was performed.

Original Research Lead hypothesis. Trade SNXUSD.HYPERLIQUID on synchronized 4-HOUR bars when Binance top-trader position-weighted long share moves opposite to the account-count long share of the SAME top-trader cohort. Follow the position-weighted change for a fixed 72 hours. This tests size concentration within a cohort, not the previously tested assertion that top traders outperform global retail accounts. Both directional arms are hypotheses, not extrapolations from successful long-side capitulation trades. Binance is a signal source only. Use market orders at the completed signal bar close; this is the registered equivalent of next-bar-open execution on a continuous venue. Protection is a platform-matched intrabar stop. No exchange-hosted protection, simultaneous opposing positions, selectable same-bar policy or latency advantage is required. Official field definitions: [Binance market-data documentation](https://developers.binance.com/en/docs/catalog/core-trading-derivatives-trading-usd-s-m-futures/api/rest-api/market-data). These definitions do not establish predictive value. Baseline signal_window_bars=180, change_hours=24 and entry_z=1.5; these signal settings are tunable under the advisory plan.

Original Research Lead hypothesis: test whether a multi-day range escape supported by unusually high same-coin spot volume persists for 72 hours on COIN-M perpetuals. Trade synchronized 4-HOUR bars; spot instruments are signal sources only. This is a conditional continuation event, not continuous momentum, spread convergence, or a mirrored capitulation fade. Market orders execute at the synchronized signal-bar close, the registered faithful equivalent of next-bar-open execution on continuous venues. Protective stops trigger intrabar in the platform; exchange-hosted protection is not required. Inverse settlement and collateral exposure must be included in account-level risk and performance. Binance describes coin-denominated margin and settlement in [its contract overview](https://www.binance.com/en-AE/support/faq/detail/85eac2bba0b342819122dc9bd4745e9b). That source supports contract mechanics, not the proposed return forecast.

Derived by the Research Lead from Aave protocol documentation and the TokenLogic buyback-pause disclosure: test whether changes in verified protocol purchases predict subsequent AAVE returns relative to ETH. The documentation states that market purchases began in April 2025 and acquired tokens enter the Ecosystem Reserve rather than being burned ([Aave documentation](https://www.aave.com/docs/ecosystem/aave)). A subsequent disclosure records a pause effective April 19, 2026 ([TokenLogic disclosure](https://governance.aave.com/t/arfc-pause-aave-buybacks/24686)). These establish a variable demand mechanism, not a published trading rule or evidence that purchases remain active today. Trade AAVEUSD.HYPERLIQUID against ETHUSDT.BYBIT using synchronized 1-HOUR bars and one daily decision. Buy acceleration in realized purchases; short withdrawal of previously material purchasing demand. Missing purchase-history data is requested below; no revenue, price, volume or treasury-transfer proxy may replace executed purchases. Market orders execute at the signal-bar close, the registered equivalent of next-bar-open execution on continuous venues. Stops are platform-matched intrabar protective stops. No preliminary performance test or qualifying-event count has been measured.

Derived by the Research Lead from Lido withdrawal-queue mechanics: test whether unusually persistent redemption demand predicts LDO underperformance relative to ETH over five days. Lido documents request locking and batch finalization in [WithdrawalQueueERC721](https://docs.lido.fi/contracts/withdrawal-queue-erc721/); that documentation supplies observable mechanics, not a trading rule or evidence of returns. Use synchronized 4-HOUR bars: short LDOUSD.HYPERLIQUID and buy ETHUSDT.BYBIT. Execute market orders at the signal-bar close, the registry equivalent of next-bar-open execution on continuous venues. No atomic cross-venue fills are assumed. Historical on-chain inputs are missing and must be acquired before implementation can test the premise. No performance pre-study was run.

Derived by the Research Lead from Kelvin J. L. Koa et al., Regime-Aware Interpretable Ranking for Sector Rotation: operationalize the section 5.1 finding that trailing twelve-month sector performance ranks below chance against subsequent one-year sector-benchmark Sharpe. This is an independently specified trading test of that finding, not a replication of RegimeRank, its Bloomberg inputs, attention architecture or reported ranking results. Use split-adjusted USEQ daily bars and annual, non-overlapping portfolio episodes. Source baseline formation and forecast horizons are approximately 52 weeks, represented by 252 trading sessions. I choose listed sector ETFs, two long and two short sectors, execution, allocation and exit rules. The platform excludes dividends, so this measures price-return reversal only. No performance pre-study was used. All market orders fill at the next session open; no same-close execution is required.

Original research hypothesis: test whether the signs of the last two non-overlapping 24-hour returns identify persistent differences in subsequent 48-hour returns, independently on four lightly tested Hyperliquid perpetuals using 4-HOUR bars. Learn four conditional means from matured historical labels with shrinkage, subtract unconditional drift, and trade only large forecasts. This is a discrete conditional-response model, not a prescribed candle pattern, continuously rebalanced panel or a claim that momentum always continues. No preliminary return study was performed; capture, frequency and performance figures below are predictions. Market orders execute at the completed signal bar close, the capability registry equivalent of next-bar-open execution on a continuous venue. Protective stops are platform-matched intrabar stops. Use all actually available history per instrument, admitting each only after its listing and complete warm-up; do not truncate every leg to the newest listing. Official base-tier perpetual taker fee reference: [Hyperliquid fees](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/fees), 0.045% per fill, 0.09% round trip. Rolling estimation is part of the frozen algorithm, not discretionary recalibration or a search on live results.

Derived by the Research Lead from ChaoZhang’s Lancelot intraday strategy, https://www.fmz.com/strategy/436765: test ordered volatility-stop agreement confirmed by a causal ATR-brick update, independently on APT, ICP and CRV Hyperliquid perpetuals using 4-HOUR bars. The complete locally archived Pine source was read. Its executable long rules, rather than the looser prose summary, define the signal. This distilled adaptation activates the source’s explicitly commented-out symmetric short rules, uses the same completed 4H clock for both components, and adds conservative sizing. These changes are Research Lead choices, not claims about the author’s tested strategy. All market orders execute at the decision bar close, the registry equivalent of next-bar-open execution on a continuous venue. Use ordinary OHLCV bars, never synthetic Renko execution prices. No preliminary performance study was performed.

Derived by the Research Lead from ChaoZhang/HPotter, Quant Strategy Based on Linear Regression Intercept: test its published signed-intercept long/short rule independently on TRX, ADA and ICP perpetuals using 4-HOUR bars. Source: https://www.fmz.com/strategy/436997 . The complete locally archived Pine source was read; the live page exposes the formula but gates the full code. Preserve the formula literally, including its negative denominator: replacing it with a conventional least-squares endpoint would test another rule. Research Lead choices are the three-name panel, four-hour interval and conservative position sizing. This is a disclosed source adaptation, not evidence of profitability. Orders execute at the signal-bar close on these continuous venues; independent opposing positions on the same instrument are never required.

Original Research Lead hypothesis: test whether traders reduce accumulated directional exposure before scheduled US CPI releases and FOMC policy statements. Independently fade each named perpetual's preceding three-day move during a roughly 24-hour window ending at least four hours BEFORE the announcement. Use 4-HOUR bars and market orders at completed signal-bar closes, the registered faithful equivalent of next-bar-open execution on a continuous venue. This is a pre-event inventory-risk experiment, distinct from previously tested post-release continuation, shock absorption and continuous cross-sectional reversal. Official calendar references: [Federal Reserve meeting calendar](https://www.federalreserve.gov/monetarypolicy/fomccalendars.htm) and [BLS CPI release schedule](https://www.bls.gov/schedule/news_release/cpi.htm). These sources establish event timing, not a trading edge. Historical release times and schedule revisions must be reconstructed from official archives before implementation; no inflation values, consensus forecasts or news sentiment are required. Estimated gross capture is an untested prediction: a 2.0% pre-event directional displacement retracing 25% yields 0.50% gross; subtract 0.09% taker round trip and an assumed 0.04% spread/slippage allowance to obtain 0.37% net. Gross capture is 3.85 times the 0.13% assumed total cost. These estimates are not measured forward returns. Predict roughly 10 qualifying events per instrument-year, or 80 trades across four names over two usable years; shared calendar dates are the independent event units, not the number of legs.

Derived by the Research Lead from Administrator / je-suis-tm, Dual Thrust backtest: test London-session range expansion on GBPUSD, EURUSD and AUDUSD using 1-HOUR bars. Source: https://github.com/je-suis-tm/quant-trading/blob/611b73f2c3f577ac5b28aaa19ac8c43d3236c7a5/Dual%20Thrust%20backtest.py . The archived full script was read. This is an explicitly causal adaptation, not a reproduction: replace its current-session rolling range, which contains future observations, with prior completed London-session ranges; use London local time with DST, hourly closes, one entry per pair per day and no reversals. Session-venue market decisions fill at the next hourly bar open. Use approximately two years of available hourly history without assuming older bars exist. No volume, swap income, queue advantage or exchange-hosted stop is required. Prediction, not measurement: an average gross directional capture of 0.30% minus a conservative 0.04% round-trip spread/commission allowance gives 0.26% before incremental impact; gross capture is 7.5 times that allowance. Actual platform commissions, minimum-order commissions, spread assumptions and impact determine realized costs.

Original research hypothesis: after unusually compressed cross-sectional returns, a sharp expansion in dispersion may mark token-specific information diffusing over several days. Trade a dollar-neutral relative-performance basket on synchronized 4-HOUR bars, not the common market direction. Baselines: formation_bars=6, history_bars=180, compression_quantile=0.30, expansion_multiple=1.8, holding_bars=18. Market orders execute at the completed signal-bar close, the registry equivalent of next-bar-open execution on a continuous venue. Intrabar protective stops are platform-matched. No queue advantage, exchange-hosted protection, selectable same-bar policy or inverse contracts are required. Use all available actual history; do not assume that Hyperliquid supplies more than approximately 5000 candles per interval. This is an unvalidated mechanism, with no preliminary performance study.

Derived by the Research Lead from OKX Learn, Sui Network and Coldware: Competing for Dominance in the DeFi Space (https://www.okx.com/learn/sui-network-coldware-defi-competition): test whether exceptional changes in Sui bridged-USDC supply precede multi-day SUI returns. The archived article was read locally; its adoption narrative supplies the mechanism, not trading rules or verified current figures. Trade SUIUSD.HYPERLIQUID long after exceptional inflows and short after exceptional outflows, using 4-HOUR bars and lagged daily on-chain observations. This requires new historical on-chain data and must wait for that dependency; price or volume must never substitute for it. Market orders execute at the completed signal bar close, the registry equivalent of next-bar-open execution on a continuous venue. Use supported platform-matched intrabar protective stops; exchange-hosted offline protection is not required. Baseline signal parameters are flow_days=3, normalization_days=60, flow_z=1.5 and min_flow_fraction=0.01. This is an unmeasured research prediction, not a claim of demonstrated alpha.

Derived by the Research Lead from the MQL5 Code Base VWAP indicator description (https://www.mql5.com/en/code/14484), read in its archived local text: test multi-day directional acceptance beyond a completed calendar month’s volume-weighted cost, independently on COMP, ZEC and BCH perpetuals using 4-HOUR bars. The source describes configurable VWAP periods, not this trading strategy. Freeze each previous UTC calendar month’s bar-volume-weighted typical price and dispersion throughout the following month. Trade newly confirmed outside-band acceptance rather than fading intraday VWAP excursions. Market entries and clock exits execute at the signal close, the registry’s faithful equivalent of next-bar-open fills on continuous venues. Protective stops trigger intrabar through platform matching. No exchange-hosted protection, independent opposing positions or selectable same-bar ordering is required. Use all available actual history; the roughly 5000-candle 4h API ceiling is about 2.3 years, not a promise of that span for each symbol. Predicted firing rate is 8–16 campaigns per asset-year; report actual distinct event dates and both arms separately.

Original research hypothesis: test temporary relative-price dislocations between the AXS and SAND gaming-token perpetuals on synchronized 4-HOUR bars. Enter a dollar-neutral pair only when its log-price ratio departs materially from its trailing equilibrium while recent return correlation remains positive. This tests sector substitution and relative-value convergence, not single-name RSI reversion, compression breakouts, funding carry or a continuously rebalanced ranking panel. Market orders at the signal bar close implement the registry equivalent of next-bar-open execution on a continuous venue. Use each instrument’s own bars and make one decision only after both bars with the same close timestamp are available; never use a later secondary bar. The pair is fixed ex ante, not selected by historical performance. Use all available per-leg history after listing and warm-up; do not backfill pre-listing observations. History length and joint event count must be reported rather than assumed. Baseline parameters are ratio_lookback_bars=180, entry_z=2.0, min_return_correlation=0.5. All return and event-count estimates below are unmeasured research predictions, not a pre-study.

Derived by the Research Lead from the prior hypothesis’s archived BigQuant summary of Huatai Securities, Smartbeta advantages in asset allocation: test a scheduled low-idiosyncratic-risk versus high-idiosyncratic-risk allocation rather than a rare dispersion event. The archived summary and linked PDF were not re-read in this revision; source provenance comes from the supplied original hypothesis. The source does not specify these trading rules. Use 4-HOUR bars on the original six-name candidate panel. Every fourteen UTC days, hold a dollar-neutral residual-volatility-ranked book for seven days, then remain flat until the next scheduled entry. The factor is constructed from the eligible traded panel, requiring no unlisted ETH signal leg. Market orders execute at the signal bar close, the registry equivalent of next-bar-open execution on continuous markets. Protective stops are platform-matched intrabar stops. No atomic basket execution or exchange-hosted protection is required.

Original research hypothesis: test asymmetric volatility-feedback episodes independently on four Hyperliquid perpetuals using synchronized 4-HOUR bars. Short when a multi-day volatility expansion accompanies falling prices; buy only when volatility contracts after a previously elevated-volatility state and prices have begun recovering. The proposed mechanism is delayed risk-budget selling during expansion and gradual risk re-entry during normalization, not a claim that an isolated crash candle must reverse. Market orders execute at the completed signal bar close, the capability-registry equivalent of next-bar-open execution on a continuous venue. Protective stops trigger intrabar in the platform; no exchange-hosted protection, hedge mode, atomic basket fills or selectable same-bar policy is required. Use all available joint history, reporting each instrument and directional arm separately. All capture figures and event rates below are untested predictions.

Original research hypothesis: test whether the direction of typical four-hour returns reasserts itself when one unusually large opposing return makes the multi-day cumulative return point the other way. Trade ETC, XLM and BCH perpetuals independently, long and short, on 4-HOUR bars; remain flat outside disagreement events. This tests robust drift versus outlier-dominated price displacement, not CUSUM persistence, swing sweeps, continuous cross-sectional ranking or a funding-carry book. Orders execute at the completed signal bar close, the capability registry equivalent to next-bar-open execution on a continuous venue. Protective stops trigger intrabar through platform matching; no exchange-hosted protection or selectable same-bar ordering is required. All capture and firing-rate numbers are unmeasured preregistered predictions. Use actual available history, at most approximately 2.3 years for Hyperliquid 4H bars, without extrapolating missing history.

Original research hypothesis: buy a completed cliff-unlock event only after the scheduled supply release and an initial price recovery, then hold seven days. Trade APT, ARB, OP, SUI and AVAX independently on 4-HOUR Binance spot bars. This tests post-distribution relief, not the prior-art pre-unlock Bybit short/BTC hedge. Historical unlock schedules, publication vintages and circulating supply are missing and explicitly requested; do not start a price-only substitute. Market orders execute at the signal-bar close, the registry equivalent of next-bar-open execution on continuous markets. Intrabar protective stops are platform-matched; no exchange-hosted protection is required. No preliminary performance study was run.

Derived by the Research Lead from kapsystem, SuperTrend Bot v6 Delta Exchange: test directional persistence of DOGE versus SHIB rather than outright coin direction. Source: https://www.tradingview.com/script/plslWgCa-SuperTrend-Bot-v6-Delta-Exchange/ (local archived full script read). Trade DOGEUSD.HYPERLIQUID against SHIBUSDT.BYBIT on synchronized 4-HOUR bars. This is an operationalization, not the author's strategy: replace candle Supertrend with an explicitly close-only relative-price version, choose the pair and venues, and replace fixed contract quantity with capped account-relative sizing. Market orders execute at the signal bar close, the registry's equivalent of next-bar-open execution on continuous venues. No independent opposing positions on one instrument, queue advantage or exchange-hosted protection is required. Dollar balance reduces common market exposure but does not guarantee beta neutrality.