Buzzberg reviewed more than 27,000 public trade ideas from over 1,700 commentators. After 20 trading days, 43.9% of the long ideas beat the S&P 500 over matching dates. That finding does not mean public commentary is useless. It means a good-sounding idea and a good investment are different things. (Buzzberg's cross-platform study, January 2–August 4, 2026)
Most equity market research guides begin with a screening ratio or a list of websites. I would begin one step earlier: What, exactly, did someone say before the stock moved? If you cannot answer that, a later chart will make almost any commentator look prescient.
The research process below is built for an investor who discovers a company through a video, X post, Reddit discussion or newsletter. Its purpose is to turn that first lead into a decision you could explain six months later, including if you were wrong.
Preserve the claim before the chart changes your memory
Save the original link and publication date. Write down the company, the direction of the idea and the reason given. A bullish mention is not necessarily an instruction to buy; a warning about valuation is not necessarily a short call. Those distinctions matter when you later judge the author and when you decide whether their reasoning applies to your own entry price.
For a usable thesis, complete this sentence: The market is missing [something] because [reason]; I will know whether this is happening when [evidence appears]. If the source only offers a ticker and a target, you may have found a lead, but you have not yet found a research case.
Keep two prices separate. The price available after the author's publication tells you how the old call played out. The price available when you discover it tells you what your own prospective return would have to beat. A writer can have been early and right while the stock is now too expensive for a new buyer.
Ask what the business must do next
Open the company's latest annual and quarterly filings. The SEC's EDGAR guide explains where to find financial statements, the business description and risk disclosures in a 10-K. You do not need to read every page in order. Start with the revenue driver named in the original thesis, then look for the figures that would prove or contradict it.
For a company that sells more units, separate unit growth from price increases. For a software company, ask whether revenue growth depends on more customers, higher spend per customer or an acquisition. For an infrastructure supplier, ask whether announced demand has become shipments and cash flow. These are different bets, even when all three companies appear in the same AI story.
Now translate the business case into what the share price requires. A useful question is not “Is this a great company?” It is “What level of future profit would make today's market value reasonable?” Build a modest case, the author's case and a case where the key assumption fails. The exercise is valuable even if you never produce a polished valuation model: it reveals which variable you are really buying.
Read the strongest objection, not the loudest objection
If the bull argues that new demand will support higher margins, the useful bear asks whether supply, competition or customer bargaining power will undo those margins. “This stock has already risen” describes the chart. It does not explain the business.
Look for disagreement in the company's own filings, competitor results and the public record. Buzzberg's trade-idea archive collects dated bullish and bearish views from the sources it tracks across X, YouTube, Reddit and newsletters. Open the original post before adopting its conclusion. The publication date, author and stated mechanism matter more than the number of people on either side.
The test is simple: can you state the bear case in a way a careful bear would recognize? If your objection is weaker than theirs, you are still arguing with a caricature. If the bear case names a measurable event, put that event next to the bull's catalyst and decide which one arrives first.
Measure the author without confusing a call with a portfolio
A stock rising after publication is not, by itself, evidence of skill. In a rising market, the relevant comparison is what the same money could have earned in a broad index over the same entry and exit dates. For an individual long stock idea, SPY is one useful reference; a sector benchmark may answer a second question about industry exposure. Choose the comparison before looking at the result.
Do not infer that the author held the stock, used your position size or exited when a model did. A recorded public idea is an observation about a statement. A portfolio return requires a separate record of trades and weights. The same caution applies to 13F holdings: the SEC allows up to 45 days after quarter-end to file, so a disclosed position is not a live view of what the manager owns today.
Our commentator study is a reminder to inspect the whole record, not just the clip that travelled furthest. In that January–August 2026 sample, fewer than half of individual long ideas beat the S&P 500 after 20 trading days. That is a result about the measured ideas and period, not a forecast for the next call.
Leave a decision that your future self can audit
Before buying, write five lines:
- The claim: the original source, date and business mechanism.
- The price: the value you are paying today and the operating result it implies.
- The opposing case: the strongest evidence that could make the thesis fail.
- The next test: the filing, earnings call or industry data that should change your mind.
- The alternative: the benchmark or other investment you would otherwise hold.
This note is short enough to finish and specific enough to embarrass you later. That is its strength. If the stock works, you can see whether your reasoning was right or whether the market rescued a weak thesis. If it fails, you can learn which assumption broke without rewriting what you believed at the start.
The best research process does not remove uncertainty. It preserves the question you were trying to answer, then lets new evidence change the answer.
Buzzberg's MCP research tools can help retrieve dated public arguments and an author's recorded history. Use that record to find the original source and the counter-case; use company filings and your own valuation work to make the decision.