How it works

Fresh Copy V1 is Pepe Capital's strategy. It follows eligible Hyperliquid wallets through new meaningful actions and independently confirmed active positions.

No. There is one active controller after migration: Fresh Copy. The old code and historical results are preserved, but its controllers cannot change the portfolio.

A new OPEN, ADD or FLIP can qualify at 10% source exposure or a 10-percentage-point increase. An existing long or short of at least 5% can qualify after two fresh, complete observations at least one minute apart. That entry is labelled CONFIRMED, not a new source trade. Smaller source positions get proportionate allocations, not an automatic 10% of Pepe.

New source actions must be at most five minutes old with a verified before/after change. Confirmed holdings instead require two independently refreshed full books within five minutes; their original trade time remains unknown. Both paths need current eligibility and executable liquidity.

Eligibility is yes or no, not a score. New entries require at least $19,000 of current source account value, positive economic history, and at least five fully observed closed trade episodes with usable entry notionals. Those complete episodes must remain profitable without the best trade. Acceptable drawdown, copyability, a clear directional book and all safety checks are also required. Missing evidence does not pass.

New capital requires a coherent research revision from the last 24 hours. An otherwise promising wallet with old evidence gets priority for a refresh within the existing research budget. Old evidence alone does not close an existing position; confirmed safety failures still do.

We use a versioned whole-book NAV receipt. Unified-account collateral is counted once from spot balances, not added again as perp equity or unrealized PnL. Standard accounts keep separate balances. Staking, vault equity and audited subaccounts are included where reconciled. Unknown account value is a data issue, not proof that a wallet is small or unskilled; unresolved debt or missing components cannot authorize new sizing.

A single unknown first hour need not erase months of measured history. We can assess the uninterrupted period after it if at least 30 days, 30 valid intervals and 99.8% time coverage are available. It is labelled observed-history evidence, not complete lifetime drawdown. Later gaps, conflicting data and excessive measured drawdown still fail.

The whole observable Hyperliquid book must be complete. A short leg might hedge spot or staking. Market makers, basis/funding strategies and portfolios whose legs cannot safely be interpreted are excluded from V1.

An eligible idea starts with up to 10% of current equity in market notional, not margin. One economic source and all its related addresses share a total 25% gross limit across every asset and both directions. Symmetric side weighting may request larger contributions, but cannot bypass that total source limit. Passing minimum checks is not proof of exceptional skill.

After an idea has been held for at least an hour, a fresh complete source portfolio can guide additional allocation within the same 25% total source budget. Larger source exposures get more of the available room. Existing positions and ordinary base targets are not cut to make the weights match. Mostly-cash or unsupported source books are not blindly magnified into a fully invested allocation. The effect on returns still needs evaluation on new observations.

No. We count independent eligible roots on each side before netting assets. Square-root weighting reduces the dominance of a much larger source count. The exact same rule favours scarce long ideas when shorts dominate. We do not cut ordinary base contributions to force balance, and do not invent the missing side. This is a sizing rule, not proof of future profits.

No. Optional allocation is reviewed at most once an hour and can add at most 50% of equity in thesis notional per review. A gap must normally reach 10% of that idea's target size, with a 0.5% equity floor and a 5% equity ceiling. Independent sources share the review budget; more coins from one source do not earn more capacity. Estimated marginal execution cost must stay within 0.15% of the addition. Source exits, stops and new qualifying entries do not wait for this review.

First by the evidence we can verify, then by active supported positions and how long ago the wallet was assessed. We reserve work for new candidates and spread research across coins and both directions. A high old profit or a big account does not automatically win priority. An old below-minimum balance can receive a limited recheck; it still cannot trade unless today's full eligibility checks pass.

Normally a single-coin breach is repaired in that coin, and a net-limit breach in the excessive direction. A linked trader portfolio is reduced together so a hedge is not accidentally removed on its own. Unrelated positions are not trimmed just to fund a new idea. Source budgets and after-cost limits remain hard constraints.

When gross exposure is below 225%, eligible active ideas may be replenished toward their approved contribution, with 250% as a refill guide rather than a mandatory exposure or a new maximum. The side rule can separately add to underrepresented ideas. No valid position is sold simply to fund another, and no stopped idea is revived. Older pre-upgrade positions need a new qualifying source ADD before optional top-ups are enabled.

Their thesis contributions are added by asset. Two independent 10% ETH longs and one 10% ETH short produce a net 10% ETH long. Each source also has its own total gross budget before netting. The controller executes only the change to the final asset target.

Yes. A short source can offset a long source in the same asset. Inspect sources to see both contributions and the resulting net market position. There is no required long/short split.

A missed quote, temporary capacity limit or expired confirmation can be checked again after 60 seconds, but only with a newer complete observation. Both confirmations must still be fresh. A stopped idea cannot reopen just because time passed.

The latest attempt stays labelled incomplete. For an existing position's 24-hour review, the last complete book can bridge the interruption only until its original five-minute validity expires. It cannot authorize buying or reset the stop. If the position must close because confirmation expires, two new complete observations after that exit can restore it.

Not as the identical old book. A new material asset or direction, or a verified new OPEN, ADD or FLIP, must establish a new source episode. Two complete observations must confirm it after the exit and, for a new action, after that action. All required legs are evaluated together within the same source and portfolio limits.

For confirmed holdings observed within one minute of each other, Pepe gives less-represented independent roots priority. Fresh economic actions retain their time priority. Entry, refill and side allocation all share the same total source budget; opening more markets cannot multiply it.

A bounded part of the research batch checks wallets with observed long and short positions. Selling alone does not mean shorting. Research coverage does not bypass wallet checks or force a trade.

No. Economic-root and cluster identities are deduplicated. Unresolved identities cannot manufacture independent consensus.

At most twelve independent source contributions per asset, 75% absolute exposure per asset, 500% total gross, net exposure between −250% and +250%, and up to 27 supported net asset positions. One economic source is limited to 25% gross thesis exposure across all markets, before netting. Limits are checked after execution costs. These are ceilings, not mandatory allocations.

At net exposure of at least +100%, eligible shorts get entry priority; below −100%, eligible longs do. Confirmed holdings have the same 5% source threshold on both sides even below those levels. Quality and source budgets are unchanged. There is no short quota and no artificial 50/50 portfolio.

Yes, when the whole observable portfolio is reconciled, the material legs are supported and the strategy is understandable. Market making, funding-dominated strategies, hidden basis hedges and missing material hedge legs do not qualify. New mixed portfolios open all required legs together or none. Later additions are planned together, not as independent directional bets.

The operating aim is 150–250% gross when enough eligible portfolios support it, not a compulsory minimum or a return forecast. Four independent sources capped at 25% each cannot provide more than 100% before netting. More genuine sources, not a higher global ceiling, are needed for broader exposure. Long and short views can remain unequal.

We net positions by asset. A 10% BTC short contribution against a 10% BTC long contribution cancels BTC exposure; it does not create two exchange positions. Both contributions still count against their sources' budgets. The source table and decision log show the two directions separately.

Yes, in either direction. Twelve independent qualifying 10% long contributions can produce about 120% exposure. The global limits have not been reduced. More independent eligible sources are needed; one wallet cannot fill the book by holding many coins. Leverage magnifies losses as well as gains, and exposure is not guaranteed.

Research-only, intraday and pair labels can trigger a closer portfolio inspection instead of an automatic rejection. Inspection is not permission to trade. A documented legacy 18% drawdown quarantine is reassessed using canonical drawdown evidence. Manual or unexplained quarantine, retirement, martingale risk and genuine safety failures are not waived.

Against fresh marks and liquidation prices in today's open perp legs, including subaccounts. A historical low buffer or an old entry price is not treated as current liquidation danger.

When the source closes or flips, after a 6.9% adverse market move from the recorded stop reference, or on an affirmative hard wallet-safety failure. A stop in one leg of a linked trader portfolio also exits its other legs. A newly proven unsupported hedge or basis structure ends copying that portfolio. Optional top-ups do not move the stop. Gaps, slippage or limited depth can delay an exit or cause a larger loss. Every 24 hours we verify the source's fresh complete book; missing evidence cannot renew a position forever. A small score change is not an exit.

New executions use the ordinary 0.045% perp taker fee, executable order-book prices and a slippage allowance. Funding uses a recent observed oracle price when available; otherwise the stored BBO proxy is labelled as an estimate. Exact exchange settlement prices are not claimed. Existing historical receipts are never repriced.

No. Copying a wallet's old fills is different from trading with Pepe's entries, position sizes and stop. We show actual Fresh Copy results separately. A replay with missing source, quote or funding history is inconclusive, not evidence of a profitable strategy.

A verified reduction of at least 10 percentage points reduces our source contribution proportionally. If the source goes from 100% exposure to 80%, we retain 80% of that contribution instead of automatically halving it. Optional top-ups cannot undo a source reduction. All contributions are netted before the final order.

After a source reduction, optional refills are disabled until a new qualifying ADD. That ADD can rebuild the base contribution and enable later side allocation, without creating another vote. Freshly verified active ideas trimmed only by portfolio limits can use bounded refills. Stopped and closed ideas still cannot reopen from unchanged information.

Not just because their last trade was more than two weeks ago. Recent qualifying research can send them for a current whole-book check. A temporary liquidation-risk veto blocks trading but does not stop monitoring an otherwise qualified wallet for recovery.

Not from unchanged information, a worker restart or a linked address. A later meaningful ADD, OPEN or FLIP can be evaluated separately. Technical confirmation failures have a separate recovery path; they are not treated as economic stop-losses. A mixed portfolio must establish a new coherent source episode before stopped legs can return.

Neither a fresh action nor a confirmed holding may have passed the checks yet. The decision log separates those paths and shows wallet, freshness, liquidity and capacity rejections. We look for eligible active exposure rather than requiring every source to trade again, but do not guarantee constant exposure.

Existing positions are closed with execution costs recorded. The previous period is archived, not rewritten. A requested capital reset starts a separate chart and return baseline. Capital added or removed is not profit; the account ledger keeps that adjustment separate from cumulative PnL.

Opening NAV plus realized price PnL plus unrealized PnL plus funding minus fees equals current NAV. Independent aggregate execution receipts reconcile to the thesis ledger. Internal netting is labelled as an internal cross, not an exchange fill.

Execution is simulated using observable quotes, liquidity, fees and funding. No real exchange orders are submitted. These results are not a verified live-trading track record or a promise of returns.