DOCS Gave Back Its Entire Pre-Market Surge — Don't Confuse a Bull Trap for a Breakout

Doximity's (NYSE: DOCS) stock exploded higher in pre-market trading this morning, surging more than 70% after CEO Jeff Tangney declared the company's clinical AI outperformed Anthropic. Then the market opened, and the rally evaporated — completely. The stock is now down 4.5% in regular trading at $20.66, having given back nearly every cent of the pre-market spike that carried it above $35.

This is not a breakout. This is a bull trap, and it's the kind of price action that separates patient investors from bag holders.

What happened

Doximity reported its fiscal first-quarter results after the close Thursday. Revenue of $156.6 million edged past the $151.8 million consensus, growing just 7% year-over-year. But adjusted earnings per share of $29 cents missed the $30 cents analysts expected. More importantly, guidance for the current quarter ($170 million to $171 million) came in below Wall Street's expectations, and full-year FY2027 guidance of $671 million to $681 million fell well short of the roughly $697 million consensus.

The stock didn't rally on its earnings. It rallied on a slide deck. Tangney pointed to a clinical AI safety study — the NOHARM benchmark, developed by researchers from Stanford and Harvard — that ranked Doximity Ask above Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol. The error rate comparison: Doximity at 4.8% versus Claude Fable 5 at 13.6%. Tangney attributed the advantage to a proprietary drug reference system and a network of more than 12,000 physician reviewers in Doximity's PeerCheck program.

Canaccord Genuity raised its price target to $36. BMO Capital lifted its to $30. The sell-side lit up. Pre-market volume spiked. And then the broader market opened, saw an earnings miss with cautious guidance underneath all the AI claims, and reversed the move entirely.

The benchmark is not what the headlines imply

The NOHARM study is real, but it's not the validation event the headlines make it sound like. It evaluated four AI systems across 1,100 simulated clinical scenarios covering 10 medical specialties, using an F1 score that balances precision and recall. Doximity Ask ranked first. But the study also found that across every single model tested, 76.6% of harmful errors were omissions — the AI leaving out critical information — rather than factual mistakes. Even the best-performing model still misses things that matter.

The study is not peer-reviewed. OpenEvidence, one of the competitors Doximity ranked above, publicly disputed the methodology, with CEO Daniel Nadler arguing that the testing format was inconsistent with rigorous medical study standards. The benchmark's authors used F1 scores before cases and answers were made public, but the scoring framework itself remains subject to interpretation.

In short, this is an industry benchmark, not a clinical trial. Tangney is a smart CEO promoting a real product advantage. But a pre-competitive safety ranking is not the same as commercial validation.

The fundamentals are solid — just not $37 solid

Here's where the contrarian in me wakes up, because the underlying numbers are genuinely compelling — just not at the price the pre-market rally briefly implied.

Doximity's gross margin sits at 89%. Operating margin is 33.3%. Free cash flow margin — the portion of every dollar of revenue that becomes actual cash after all operating expenses and capital investments — is 49.2%. That's software-industry elite. The company carries $273.6 million in cash against only $167.8 million in total debt, with a debt-to-equity ratio essentially at zero. Return on invested capital is 13.9%, return on equity 17.2%.

Revenue growth of 11.2% year-over-year isn't spectacular, but it's not a failure either. The stock trades at a forward P/E of roughly 17.3x and a trailing P/E of 22x. Enterprise value to EBITDA is 14.2x. For a company with 89% gross margins, 49% free cash flow conversion, and virtually no debt, the market is pricing it like a sluggish legacy software name rather than a high-margin platform.

This is the kind of valuation compression that usually catches my attention. A 17x forward P/E on a business growing 11% with margins this thick — the market has arguably baked in a scenario where pharma advertising never recovers and AI investments destroy profitability. That doomsday narrative doesn't match the numbers.

But the timing is terrible

DOCS is down 69% over the past 12 months. Year-to-date it's down 53%. The 52-week range stretches from a low of $17.15 to a high of $76.51. The stock is below both its 50-day moving average of $21.14 and its 200-day moving average of $32.58. The RSI at 44 is in neutral-to-weak territory. The MACD line is negative. The 1-day volatility is 12.9%, meaning the stock is whipsawing violently.

The pre-market-to-regular-trading reversal — from a 70%+ spike to a 4.5% decline — is the most telling data point. It tells me the initial buyers were short squeezers and momentum chasers who front-ran a narrative, not institutions building positions on fundamentals. When the opening bell rang and real money assessed an earnings miss with weak guidance, they didn't buy the pop. They sold into it.

AInvest's aggregate signal rates DOCS as a Hold with a composite analysis rating of 0.92, despite giving it a near-perfect 9.88 on fundamentals. That disconnect — stellar fundamentals, tepid overall rating — captures the essence of the situation: the business is in good shape, but the near-term setup is messy.

The real question: falling knife or falling opportunity?

I'm not dismissing the longer-term case. Doximity controls a network of roughly 8 million U.S. healthcare professionals — that's a moat that's hard to replicate. The PeerCheck program, with thousands of physicians reviewing AI outputs, gives the company a quality layer that pure-play AI firms don't have. AI monetization isn't here yet — management called FY2027 an "AI investment year," which means near-term margin pressure — but the pipeline from pharma search marketing and enterprise Ask contracts could materialize as Canaccord forecasts in Q3 and beyond.

But the market doesn't pay you for being early. And it certainly doesn't pay you for chasing a 70% pre-market spike that dies at the open.

I don't think investors need to chase this rally. The fundamental setup is attractive at current levels — a 17x forward P/E on a high-margin, cash-generative, debt-free business is arguably dirt cheap relative to the growth profile. But the better entry is not today's carnage. It's after the dust settles, after price action shows selling exhaustion rather than another reversal, and after management proves that AI isn't just a margin-draining investment thesis but a revenue driver.

I would reassess the bullish case if next quarter's guidance drops further below consensus, if gross margins compress below 80% as compute costs accelerate, or if the pharma advertising business shows structural deterioration rather than a cyclical pause. Those are the moat-cracking risks, not a disputed benchmark study.

For now, DOCS is a stock I'd watch on my list, not one I'd buy in the panic. The fundamentals deserve patience, not a chasing trade.