Phase Space AI

Trade Construction

Doximity [DOCS]

Doximity [DOCS] — Trade Construction & Liquidity Criteria

No position verdict is issued here. This section establishes how a position would be expressed and whether it can be, if the book elects one.

1. Liquidity Criteria — PASS (equity), FAIL (options at size)

Equity liquidity is ample for any size this book would take.

2. Vehicle ladder — rung 1, EQUITY

Chosen vehicle: common equity. Ladder position 1. No argument required (criteria.md: "Equity is the default and needs no argument").

Required disclosure, given whatever the vehicle

Implied volatility minus trailing realised volatility, as a number:

Quoted size at the specific strike (not chain-level open interest, per §"Depth of position is not depth of quote"):

Why not the LEAP, explicitly

A Jan-2027 LEAP is 5.5 months out — shorter than the declared 12-month thesis horizon, which puts it on rung 3 and requires both an argument and a catalyst before expiry. The only expiries Alpaca returns beyond 2027-01 are 2027-02-19 (9 call contracts, max OI 19) and nothing further. So:

• There is no expiry that is both liquid and long enough. Per criteria.md's horizon-conflict rule, "Where no expiry is both liquid and long enough, the answer is equity."
• Even setting horizon aside, one would pay a +9.0 point implied-over-realised spread, on a name whose realised vol is already 57%.
• The bid-ask at the ATM strike is 27.4% of mid. Round-tripping costs ~27% of premium before any move.
• Multi-leg (rung 4) is unconstructable on the same grounds: the 2027-02-19 chain's maximum open interest across all 9 contracts is 19, reproducing the FCEL failure mode exactly (12-contract short leg).

Conclusion: equity, with no derivative overlay. This is the same conclusion every chain pulled in this project on 2026-07-29 reached, and for the same measured reason.

3. Position construction, if the book elects one

The Valuation Criteria returns FAIL, so the long case is not established by this memo and no entry is proposed. Recorded for completeness if the book takes a view against this memo's conclusion:

• Entry discipline: would require evidence that NRR has stabilised at or above 109% in the Q1-FY27 print — the single disclosure that would falsify the deceleration thesis. NRR is disclosed annually in the 10-K, not quarterly, which means this thesis cannot be monitored at quarterly frequency from the KPI itself. Use revenue growth versus the +3.9% guide as the quarterly proxy.
• Invalidation: a print above ~+8% YoY revenue growth, or a first disclosed dollar figure for the Clinical AI Suite, would materially change the required-versus-demonstrated arithmetic.
• Sizing input: inverse-volatility at 57% realised vol places this name in the small bucket. Note the known limitation — HALO demonstrated that inverse-vol sizing misprices a fat left tail; here the left tail is cushioned by $748.6m of net cash (18.9% of market cap) and no debt, so the usual concern is less acute than the vol number implies.
• Exposure tags for the book: sub-sector HCIT; correlates with HIMS (same tag). Concentration check against HIMS required before any addition.

4. Short Mechanism Criteria (MEASURED — scored, acted on by nothing on this fork)

Both legs are present and this is recorded, not acted on: - Decelerating growth: 17.0% (4y) → 15.5% (3y) → 13.1% (1y) → +3.9% guided. - Exhausted margin runway: GAAP operating margin 33.3%, already peaked at 39.9% (FY2025), now contracting; adjusted EBITDA guided down 6–10%; SBC at 18.9% of revenue and rising. Incremental leverage visibly spent.

Also noted per §9 of known-silent-failures: a negative long expected return is not a positive short expected return. Any short here would need borrow, slippage and the $748.6m net-cash floor priced on their own before the sign could be trusted. Not done, because this is a long-only book and the fork is not live.