Read a Trader's Thesis Before You Copy Their Trade

How do I read a trader's thesis before copying their trade?

Read the words second. First, find out whether the person who wrote them still owns the coin, how they got it, and what the chain says about their results.

A thesis is a public claim. It costs nothing to write, it can collect likes from followers who never bought, and it is posted after the position exists. The wallet is the evidence. If the two disagree, trust the wallet.

Here is the prompt we gave Claude, then the exact call it made.

Show me GL_AW_AT_PARAGON's most recent theses. For each one, tell me
the PnL on that coin and whether this trader still holds it.

stalkchain_fomo_theses_by_trader
  trader: "GL_AW_AT_PARAGON"
  limit: 3

Trimmed response:

{
  "data": {
    "key": "user:gl_aw_at_paragon",
    "kind": "theses",
    "capturedAt": "2026-10-08T06:13:45.836Z",
    "theses": [
      {
        "tradeId": "25375580-0ee7-4ace-ab8a-e914b0bcf1fc",
        "text": "SpaceXSi.com sold for $174,000 USD via Spaceship to @elonmusk",
        "handle": "gl_aw_at_paragon",
        "ts": "2026-10-07T17:17:50.000Z",
        "likes": 69,
        "replies": 0,
        "tradeUsd": 222031.19,
        "equity": null,
        "realizedPnlUsd": 7689.22,
        "unrealizedPnlUsd": 123433.49,
        "token": { "symbol": "SI", "address": "DEW9…WDP" },
        "chain": "solana"
      },
      {
        "tradeId": "6afbde6a-c71a-4094-884d-40f3a6a878af",
        "text": "Now Playing: Linkin Park - Crawling",
        "ts": "2026-10-06T00:42:25.000Z",
        "likes": 35,
        "tradeUsd": 53254.47,
        "equity": null,
        "realizedPnlUsd": 0,
        "unrealizedPnlUsd": -10365.81,
        "token": { "symbol": "CRAWL", "address": "BXoH…pump" }
      }
    ],
    "count": 2
  }
}

Pulled Oct 8, 2026, 06:13 UTC via StalkChain's MCP (stalkchain_fomo_theses_by_trader). Fields and tokens are trimmed, and the token addresses are shortened.

Same call over REST:

curl -G https://data.stalkchain.com/api/v1/tools/stalkchain_fomo_theses_by_trader \
  -H "Authorization: Bearer sc_YOUR_API_KEY" \
  -d trader=GL_AW_AT_PARAGON \
  -d limit=3

GL_AW_AT_PARAGON is a public FOMO handle. We use it as an example of how to read the data, not as a verdict on the person.

Is this handle's wallet real, and who says so?

Before you weigh any words, confirm which wallet belongs to the author. A thesis tied to the wrong wallet gives you a clean-looking, wrong answer.

stalkchain_fomo_search
  q: "GL_AW_AT_PARAGON"
  type: "traders"

{
  "data": {
    "count": 1,
    "results": [
      {
        "type": "trader",
        "handle": "GL_AW_AT_PARAGON",
        "userId": "85fb7b04-fdb3-5e2c-8725-a6f8f026cc76",
        "wallets": {
          "solana": "3nsE…vQ1B",
          "evm": "0x0ecb…4c2b",
          "verified": true,
          "evidence": { "solana": "onchain-holdings", "evm": "onchain-holdings" }
        },
        "source": "directory"
      }
    ]
  }
}

The field that matters is evidence. onchain-holdings means the wallet matched what the trader holds on chain. The other value you can see is provider-claimed, which is a lead, not proof. Treat a claimed wallet as unverified and do not run a PnL cross-check on it as if it were the author's.

Here both wallets carry the stronger flag, so the checks below are fair game.

What does each field in a thesis mean?

Read the thesis payload field by field. Three of them are easy to misread.

The tradeUsd caveat is the big one. A reader sees "$222,031" next to a thesis and assumes the trader put in $222k. When we pulled the underlying trade, the real fills were 7 USDC buys between Oct 4 and Oct 6, from $85 to $4,108 each, adding up to $6,092.

Of the 15.3M SI tokens in the position, about 13.4M arrived by transfer, not by buying. The cost basis counts those tokens, and so does the unrealized profit. The pull of that trade lookup also noted that the upstream trade endpoint answered with a 503, so the figures were assembled from positions and swaps instead.

Why did limit=3 return only 2 theses?

We asked for three and got two. The response says count: 2, and it does not say whether that is all the trader has ever written.

It is not. A separate call, stalkchain_fomo_trade_comments, shows five theses on the SI trade alone. The newest one reports olderThesisCount: 8, which means at least nine theses on that coin. So count here is "returned in this window", not a lifetime total.

We do not know the cause. It may be a recency window or a cap on the endpoint. Whatever it is, do not read "2 theses" as "this trader rarely explains their trades."

To get the rest, narrow the query instead of raising the limit. The tool accepts tokenAddress, minLikes, and sort, so you can ask for one coin or for the most-liked posts. The next section shows why this matters.

Which of this trader's theses did the platform rank best?

The newest thesis is not the best one. FOMO picks a "best thesis" per trader, and stalkchain_fomo_trader_spotlight returns it.

stalkchain_fomo_trader_spotlight
  trader: "85fb7b04-fdb3-5e2c-8725-a6f8f026cc76"

{
  "data": {
    "bestTrades": [
      {
        "token": "SI",
        "costBasisUsd": 222031.19,
        "realizedPnlUsd": 7689.22,
        "unrealizedPnlUsd": 123433.49,
        "openedAt": "2026-09-24T04:38:40.000Z",
        "thesisLikes": 69
      },
      {
        "token": "e/acc",
        "costBasisUsd": 16057.78,
        "realizedPnlUsd": 1838.23,
        "openedAt": "2026-09-26T02:30:55.000Z",
        "closedAt": "2026-09-26T08:20:05.000Z"
      }
    ],
    "bestTheses": [{ "token": "SI", "thesisLikes": 453 }]
  }
}

Pulled Oct 8, 2026, about 06:14 UTC via StalkChain's MCP (stalkchain_fomo_trader_spotlight). Thesis text is trimmed.

Two things stand out. The best-rated thesis has 453 likes, not the 69 on the newest one, and both belong to the same SI trade. So there are only two distinct trades here, not three.

Also notice the dates. The spotlight says the SI position opened on Sep 24. The USDC buys we found were on Oct 4 to Oct 6.

Those two facts fit together if the early part of the position arrived by transfer. The payload does not say so, so treat it as a question to ask, not a finding.

The e/acc row is the cleaner example. It opened and closed within about six hours and booked $1,838 realized. That is a trade you can evaluate without guessing about transfers.

Does the author still hold the coin they wrote about?

Yes, in this case, and heavily. This is the check that separates a thesis from an exit.

stalkchain_fomo_trader_balances
  trader: "85fb7b04-fdb3-5e2c-8725-a6f8f026cc76"
  limit: 5

{
  "data": {
    "totalValueUsd": 376593.92,
    "holdings": [
      { "symbol": "SI", "chain": "solana", "amount": 15330944.86,
        "priceUsd": 0.0216563, "valueUsd": 332011.68, "change24h": -25.2197 },
      { "symbol": "CRAWL", "amount": 10880352.6, "valueUsd": 43356.38, "change24h": 40.4385 },
      { "symbol": "DARK", "valueUsd": 1192.02 }
    ],
    "byChain": { "solana": { "holdings": 7, "valueUsd": 376593.91 } },
    "_truncated": { "returned": 5, "total": 8 }
  }
}

Pulled Oct 8, 2026, about 06:14 UTC via StalkChain's MCP (stalkchain_fomo_trader_balances). We set limit: 5, so three of eight holdings were cut, and _truncated says so.

SI is $332,012 of a $376,594 portfolio, about 88%. The author wrote the thesis on Oct 7 and was still holding more than 15M tokens the next morning, with the coin down 25% on the day.

Someone who posts a thesis and sells would show a shrinking balance or a realized figure far larger than $7,689.

A concentrated holder is aligned with the thesis. It also means one coin decides their results, and any exit you copy later will move that price.

Does the platform PnL match the on-chain PnL?

Not cleanly, and that disagreement is the most useful part of the exercise. We ran the same Solana wallet through a raw on-chain PnL check, stalkchain_wallet_pnl, which does not depend on FOMO's position model.

Pulled Oct 8, 2026, 06:00 to 06:20 UTC via StalkChain's MCP (stalkchain_fomo_trader_report, stalkchain_wallet_pnl, stalkchain_fomo_trader_balances).

The two columns use different methods and different samples. The FOMO side looks at one coin and a sample of 25 closed positions, and it counts transferred-in tokens at a cost basis that makes them look like profit.

The on-chain side looks at the wallet's whole traded history. The roughly $6,700 gap in holdings is most likely price timing.

We are not saying which column is right. We are saying that a reader who only sees the +$123k next to a 69-like thesis has half the picture. The big green number depends on how the tokens arrived.

For a deeper walk through those win-rate numbers, see what counts as a good wallet win rate and how to check whether a Solana wallet is profitable.

How should I weigh a thesis? Good patterns and risky ones

Treat the thesis as a hypothesis and let the wallet grade it. These are the patterns we look for.

None of these is a verdict on its own. A transfer-in can be a trader moving tokens between their own wallets. A thin sample can just mean a young account. The point is to know which questions your copy decision rests on.

What should I not conclude from this?

Do not conclude that the trader is dishonest, that the coin is a scam, or that the thesis is wrong. The data supports none of those.

It supports narrower statements. The headline PnL on this thesis leans on tokens that arrived by transfer. The on-chain PnL for the wallet is negative. The author is concentrated, and they still held with the coin down 25% on the day.

Also avoid treating likes as a signal of future price. 453 likes on an older SI thesis tells you the post spread. It tells you nothing about the next move. Data, not financial advice.

How does this compare with other thesis data?

FOMO is the origin of this data, and its API publishes a thesis data feed explainer and an opinion piece on why thesis data matters. As of October 2026 we could not find another provider that serves written theses as data. That makes it the right source, and it is why this article exists.

What those pages do not show is a thesis sitting next to the author's on-chain record.

The checks here are what we add: the evidence flag on the wallet, the cost-basis versus fills comparison, the independent on-chain PnL, and the current balance. Together they let you ask whether the author still holds, which is the question a like count cannot answer.

Nansen has started labeling wallets with FOMO handles, but as a label inside its own tools, not as thesis text. For the fuller comparison, read StalkChain vs Nansen.

How do I run this check myself?

Connect StalkChain's MCP to your AI client and ask the question in plain English. The Claude setup for FOMO data takes a few minutes, and the same data is available through the StalkChain trader data API and the MCP server.

A prompt that does the whole workflow:

For the FOMO trader GL_AW_AT_PARAGON: confirm the wallet evidence flag,
list their latest theses with PnL, check whether they still hold each
coin, and compare the platform PnL with the on-chain PnL. Flag any
transferred-in tokens.

Theses by trader also cover Robinhood Chain, so the same method works on Robinhood Chain data. Solana forensics tools such as deployer and sniper checks stay Solana-only.

Data, not financial advice. These are examples of how to read a payload, not signals to buy or sell anything.

FAQ

How do I check a trader's thesis before copy trading?

Resolve their handle to a wallet, read their latest theses, then compare each one with their cost basis, their on-chain PnL, and their current balance. If the author no longer holds the coin, or if most of the position arrived by transfer, discount the thesis.

Do likes on a thesis mean the trade was good?

No. Likes measure how far a post spread among FOMO users. In our pull, the newest SI thesis had 69 likes and the best-rated one had 453, and neither number says anything about entry price, size, or outcome.

Why does the thesis PnL differ from the on-chain PnL?

They use different methods and samples. The thesis figure is tied to one position and the platform's cost basis, which can count transferred-in tokens. The on-chain check measures the wallet's whole traded history. When they disagree, find out how the tokens arrived before trusting either.

Why did the theses endpoint return fewer results than my limit?

We asked for three and got two. The same trader has at least nine theses on SI alone, per the trade comments lookup, so the count is not a lifetime total. Narrow the query by token, minimum likes, or sort order instead of assuming the history is short.

How can I tell if a trader still holds the coin they wrote about?

Check their live balances for the thesis token. Compare the amount with the position size and with the date of the post. A balance that holds or grows after the post is consistent with conviction. A balance that falls as likes rise is worth a closer look.

What does provider-claimed mean on a trader's wallet?

It means a source said the wallet belongs to the trader, but the wallet did not match the trader's on-chain holdings. Treat it as a lead. The stronger flag is onchain-holdings, which means the holdings matched.