Harvest.art

How To Run A Monthly Crypto Review

Farmer Harvey

You do not need a new edge. You need a calendar. One standing hour a month beats heroic triage in December. Think of it as lightweight compliance for your future self: you write down what you own, decide whether you still believe it, and if not, you clean up the mess. That’s the whole program.

This is not prediction, or connoisseurship, or macro. It is housekeeping. You are inventorying claims you made in the past (by buying them) and deciding if you’d make them again (by keeping them). The byproduct is fewer unforced errors, tidier tax outcomes, and slightly more sleep.

The agenda

  • Inventory: holdings by wallet, USD cost basis per lot, current bids/floors, and position sizes as a percent of your total stack. If you can’t see it in one page, it doesn’t exist.
  • Health check: liquidity, spread, buyers per day, turnover. Not vibes; microstructure. If you must use vibes, put a number on the vibe.
  • Actions: realize losses, rotate into a close substitute you prefer, or explicitly do nothing. “Do nothing” counts if you write it down.

The rules

  • Below basis, no love, no problem: if a lot is below basis and you would not buy it today, sell it and move on. You are not obligated to hold your past.
  • Still believe, switch vehicles: if you like the exposure but not the instrument, rotate into a close substitute you prefer (tighter spreads, deeper liquidity, cleaner execution). Same idea, better wrapper.
  • If the thesis changed, say it out loud: one paragraph: what you believed, what changed, what you’re doing. Pretend you are explaining it to future-you, who will be weirdly judgmental.

The outputs

  • One-page log: a dated note listing changes, the why, and the numbers. No poetry.
  • Tagged transactions: tx hashes, receipts, and screenshots in the same folder as the note. Your accountant is a person; be kind to them.
  • Sizes aligned with risk: after changes, your top 5 positions should fit the risk budget you actually intend, not the one inertia delivered.

A 45-minute script

  • Minutes 0–10: Snapshot

    • Export holdings by wallet. Include USD cost basis per lot, not just per position. List current bids/floors and spread for each top holding. Note unrealized PnL in USD by lot.
    • Tag your top exposures by theme: “ETH beta,” “marketplace tokens,” “PFP midcaps,” “stables.” Percent-of-book next to each.

  • Minutes 10–25: Candidates and substitutes

    • Circle lots below basis that you would not buy today. If you wouldn’t add at this price, why are you holding it at this price?
    • For each circled item, propose a substitute that serves the same thesis but with better microstructure: tighter spread, more buyers, clearer catalysts, less headline risk. Write the pair: “From A → Into B.”
    • Stage orders: set a realistic target to sell (often the top bid) and a resting bid for the substitute below the mid, sized to your plan.

  • Minutes 25–35: Execute

    • Fill the sales first. Realize losses on the circled lots. Do not negotiate with yourself while the transaction is pending.
    • Place resting bids for the substitutes. You don’t have to be a hero; good enough is good enough. If they don’t fill, you already did the important part.

  • Minutes 35–45: Log and resize

    • Record the actions: what you sold, size, price, realized PnL per lot, substitutes and bid levels. Paste tx hashes. Screenshot receipts. Save to your review folder.
    • Update your one-paragraph theses. Adjust position sizes to your intended risk budget. If you did nothing, say “I did nothing” and why.

Example: rotating an NFT

  • You own Collection A #4321 with a recorded USD cost basis of $5,400. Floor is 1.2 ETH, top bid 1.17 ETH, spread 2.5%, buyers/day ~35, turnover ~0.8%.
  • You still like the category (cartoon animals that are, somehow, also a chain of diners) but prefer Collection B: similar vibe, spread 3%, buyers/day ~120, turnover 2%.
  • Action: take the 1.17 ETH top bid on A, realize a USD loss on that lot based on your recorded basis, and rest a 0.85 ETH bid on B at mid minus spread. If it fills, great; if not, you still improved your tax lot and cleaned up the book.

Example: rotating a token

  • You own Token X with a $2,400 USD basis, it trades 1,820/1,835, and the order book is a wasteland on weekends.
  • You want “Layer 2 beta” more generally, and Token Y (an index-y wrapper) trades 0.5% spread, deep markets, and 24/7 liquidity.
  • Action: sell X at market, realize the loss, set a ladder of bids for Y. Same thesis, better execution environment.

Why this works (and why it’s boring on purpose)

  • Microstructure > narrative: when you are cleaning a book, your adversary is friction: wide spreads, illiquidity, fees, and your own reluctance. Liquidity, turnover, and buyers/day are leading indicators of whether you can get out later when you need to.
  • Taxes are part of returns: realized losses can offset other gains. As of this writing in many jurisdictions there is no explicit “wash sale” rule for crypto. You should still behave like an adult: do not play games, document intent, and talk to an actual tax professional if the numbers are meaningful to you.
  • Behavioral hygiene: writing “keep this because sunk costs” feels bad on paper, which is why the paper is useful. The ritual is a nudge to make explicit choices instead of implicit drift.

Practical setup

  • Calendar invite: recurring, last business day of the month, 45 minutes, with a link to your review folder.
  • One-page dashboard: wallets, lots, USD basis, current bids/floors, unrealized PnL in USD by lot, spread, buyers/day. If you need to scroll, it’s too long.
  • Record basis correctly: capture USD cost basis at fill time for every lot using the transaction’s fiat reference price; do not track basis in token units.
  • Templates: a short note template and a folder structure: Reviews/YYYY/MM/. Name files YYYY-MM Review.md and YYYY-MM Receipts/ for screenshots and PDFs.
  • Tags: in your tracker, tag each action with a verb: Sell, Rotate, Hold. Future-you can filter by quarter and see what you actually did.

Common objections

  • “What if it rips the day after I sell?” Then it rips. Your process is not a prophecy; it is a broom. Clean the floor again next month.
  • “Gas and fees add up.” True. They also add up when you panic in December. The monthly cadence spreads the cost and usually gets better execution.
  • “This is too rigid; I’m a long-term holder.” Great. The review can end with “do nothing” for months in a row. The value is that you meant to do nothing.
  • “What if there is no good substitute?” Then reduce exposure and wait. Cash is a position; clarity is a substitute for most things.

A tiny template you can paste

  • Date: YYYY-MM-DD
  • Portfolio value: X; Top 5 exposures (%): A/B/C/D/E
  • Changes:
    • From: Asset/Lot, Size, Price, Realized PnL
    • Into: Substitute, Target Size, Bids
  • One-paragraph thesis updates: [Short bullet]
  • Receipts: [Links/hashes]

Consistency is a form of alpha when everyone else improvises. Ship the calendar invite. Write the one page. Do it again next month.