How to Use Perplexity for Stock Research: 2026 Guide

Open Perplexity Pro, click the focus selector, and pick Finance. This changes the source mix to weight SEC filings, IR pages, earnings transcripts, sell-side notes, and credible financial journalism above general web. Next, create a new Space named exactly with the ticker symbol you are researching. Spaces persist context across queries, so every subsequent search on the ticker builds on the prior ones. The combination of Finance focus plus a dedicated Space is the single biggest setup decision that compounds across a full workup. Without Finance focus, Perplexity returns more general web results and you will spend time filtering. Without a Space, every query starts cold and you lose the running context that makes the third and fourth queries on the ticker materially better than the first. For active investors with more than 5 to 10 positions under research, build a system: a top-level Space for sector or theme, sub-Spaces for individual tickers, and a master research journal where you periodically export the highlights. The 10 minutes of setup pays back inside the first earnings cycle.

Drop the 10-Q or 10-K PDF into your Space (or paste the EDGAR URL) and ask 6 to 8 structured questions rather than reading the filing linearly. The questions that produce the most insight per minute: what are the 10 most material risks named in Item 1A and how have they changed vs the prior period; what is the full revenue breakdown by segment and geography for the last 3 years; what are the 5 most material changes in the MD&A vs the prior period; what off-balance-sheet liabilities or contingencies are disclosed anywhere in the document; what is the auditor's status (any change in auditor, restatement, or going-concern language); what is the working capital trend (DSO, DPO, days of inventory) over the last 4 quarters; what is the capex trajectory and the breakdown by category if disclosed; what is the executive compensation structure including the performance metrics and any clawback provisions. Each question returns a focused 100 to 300 word answer with citations to specific pages of the filing. This compresses a 200-page filing to a 30-minute structured read with no loss of material content; the only thing missed is the second-order observations that come from human pattern-matching across years, which is where your own follow-up reading goes.

Drop the most recent earnings call transcript into the Space and ask 5 structured questions that catch what linear reading misses. First, list every unprompted topic where management raised something the analysts did not ask about (these are often the things management most wants to communicate). Second, identify every walk-back or softening of prior guidance, with the prior language quoted and the new language quoted side by side. Third, list every new metric introduced (companies often pre-announce a new KPI as a leading indicator before formalizing it). Fourth, identify every analyst question that visibly made management uncomfortable, where the answer was qualified, redirected, or unusually long; rank these by how material the underlying topic appears to be. Fifth, ask for the macro environment commentary changes vs the prior quarter (was demand described as resilient, mixed, weakening). These 5 questions extract more signal from the transcript in 15 minutes than a full linear read produces in 60 minutes. Combine with the consensus estimate trajectory query from step 5 to see whether the surprises in the call have been priced in by the sell side.

Ask Perplexity for all Form 4 insider transactions for the ticker in the last 12 months. The answer should include the names and titles of the insiders, the transaction dates, the transaction type (open market purchase, planned sale under 10b5-1, exercise and hold, exercise and sell), the share counts, and the dollar amounts. Then ask Perplexity to identify the pattern: is the buying clustered or scattered, is the selling 10b5-1 (mostly noise) or discretionary (often signal), are the largest transactions from C-suite or board members, has the pattern changed in the last 90 days vs the prior period. The interpretation that holds up over time: insider buying by multiple insiders at market prices, especially clustered in C-suite and board, is a moderately bullish signal that the sell side often misses for a few weeks. Insider selling is mostly noise unless it is unusually large, clustered in time, or breaks an executive's historical pattern. The cases where insider selling is signal: discretionary sells right before bad news (visible only retroactively), unusual concentration of multiple insiders selling in a short window, executives selling at a pace materially higher than their historical baseline.

Ask Perplexity for the consensus estimate trajectory over the last 90 days for revenue and EPS in the current quarter, next quarter, and current year. Ask which direction the trajectory is moving and whether the latest estimates are above or below the company's most recent guidance. Then ask for the top 3 sell-side notes published in the last 30 days summarized in 3 sentences each, with the analyst name, the firm, the price target, and the key disagreement vs consensus. Then ask for the bear case from credible critics, specifically requesting short-seller reports, skeptical analyst notes (the analyst who has been bearish and is sticking with it), and any recent activist or short-interest disclosure. The pattern that produces good decisions: when the company narrative, the consensus trajectory, and the price action all move in the same direction, the market is digesting the story consistently. When they diverge (company guides higher, consensus falls, price drops), there is a thesis-shaping mismatch to investigate. Use the divergences as the starting point for deeper work, not the agreement.

Build a side-by-side comparison of the subject ticker against 3 to 5 peers plus the sector ETF or relevant benchmark. The metrics that produce useful comparisons: market cap; revenue growth (last year and 5-year CAGR); gross, operating, and net margin; free cash flow margin; return on invested capital; net debt to EBITDA; forward P/E; EV/EBITDA; EV/Sales; dividend yield; 1-year and 5-year total return. Ask Perplexity to identify the 3 metrics where your subject ticker is the cleanest outlier (positive or negative) and to surface the most recent piece of credible analysis explaining each outlier. The reason peer comparison matters more than absolute analysis: a stock can look attractive in isolation but be average vs peers, in which case the entire sector is cheap and your name is not the best expression. Or it can look expensive in isolation but cheap vs peers given quality differences. The 3-metric outlier framing is what produces actionable thesis points: if your ticker is the cleanest positive outlier on free cash flow margin and ROIC, the thesis writes itself; if it is the cleanest negative outlier on revenue growth, the thesis better explain why that is improving.

Most retail investors under-research the bear case because it is uncomfortable to imagine being wrong. The fix is structured: ask Perplexity for the most-cited bear case on the ticker in the last 90 days; the current short interest as a percentage of float and how it has trended in the last 90 days; any recent short-seller reports (Hindenburg, Muddy Waters, Citron, others) with publication dates and the core allegations; any pending litigation, regulatory action, or SEC investigation. Then ask Perplexity to steelman the bear case: regardless of your own view, write the strongest possible argument for shorting the stock at the current price, using only credible cited sources. Then write your own response to each point in the bear case, in your own words, citing your own evidence. The exercise of writing the rebuttal yourself (not letting Perplexity do it) is what hardens the thesis. If you cannot rebut a specific bear point with cited evidence, that is a real gap in the thesis and the position size should reflect it. The investor who runs this exercise on every position will be wrong less often than the investor who does not.

The final step is the one most investors skip and the one that makes all the prior steps worth doing. Open a blank document and write a one-page thesis in your own words. Required sections: the company in 3 sentences; the bull case in 5 bullets; the bear case in 5 bullets; the catalyst calendar over the next 6 to 12 months; your specific entry and exit criteria with price levels; your position sizing logic. Use only the cited sources from your Perplexity Space; do not paste Perplexity output. The discipline of writing the conclusion yourself is what makes you the investor and Perplexity the research analyst, not the other way around. The reasons this works: writing forces you to commit to specific predictions that you can grade later; writing exposes gaps in the thesis where you do not actually have a view; writing creates the journal artifact that becomes the most useful asset in your investing process 6 to 24 months later. Save the thesis in the Space. Update it after every quarter, every major event, and every price move of more than 15%. Six months later, the comparison between what you wrote then and what you think now is the single highest-leverage learning loop in investing.