ServiceTitan [TTAN] — Valuation

As of 2026-07-29 · spot $78.40 (2026-07-28 close) · framework v1.5.1 Two outputs are produced, per references/valuation.md. Neither replaces the other.

0. Corrected inputs — everything below rests on these

The screen's figures ($2.775bn cap, 2.42x EV/S, $317.5m net cash) are wrong; see TTAN_Research.md §0.

Demonstrated growth — I do not adopt the screen's 27.1%

I use +24.1% as demonstrated — the most current full-period measure, and the most conservative of the three I can derive from filings. Using the screen's 27.1% would add 3.0pp to every margin below.

1. The implied-path test — the Valuation Criteria (PRIMARY output)

Reverse DCF, assets/reverse_dcf.py. Terminal value is 100% of EV by construction in this instrument, so per criteria.md this is mandatory as the primary long-horizon output and a forward DCF is not run as a verdict.

Parameters

Solved for: 5-year revenue CAGR. Held fixed, named explicitly:

1.1 The result

THE MARKET REQUIRES A 19.8% 5-YEAR REVENUE CAGR

Demonstrated: 24.1% · Required: 19.8% · MARGIN = +4.3pp

Valuation Criteria: PASS — but thin, and thinning

The implied path sits below what the business has already demonstrated, so this is a PASS on the stated test, not a PASS-WITH-ARGUMENT. No narrative is required to reach it.

But the honest reading is that the cushion is nearly gone. Company guidance for FY2027 is +18.1% at the midpoint — below the 19.8% the price requires. My own FY27 estimate of $1,190m (+23.8%, built off the demonstrated beat pattern in §3) clears it, but only by 4pp, and the deceleration is real: +25.6% (FY25) → +24.5% (FY26) → +23.8% (FY27E, house) → guide +18.1%. The required path and the actual path are converging. This is a PASS whose margin is a function of one more year of execution, not a structural cushion.

1.2 Terminal EBIT margin of 20% — built, not assumed

The screen used 11.2%, described as "industry median (pre-profit)" — a generic placeholder that ignores a 78.7% platform gross margin. I rebuild it from the disclosed structure:

SBC is included, not added back. TTAN's SBC is 20.5% of revenue today, of which ~5.2pp is the Co-Founder market-condition award that expires October 2034 and rolls off. Ex-Co-Founder SBC is ~15.3% of revenue; mature software runs 8–12%. The 20% terminal margin embeds SBC normalising to ~10% of revenue. If SBC does not normalise, the terminal margin is ~15% and the name fails — see the grid.

Cross-check against the company's own trajectory. Non-GAAP operating margin (which excludes all SBC, acquired-intangible amortisation and lease losses): 3.3% (FY25) → 9.8% (FY26) → 12.7% guided FY27 → 15.2% already achieved in Q1 FY27. A 20% GAAP-inclusive-of-SBC terminal margin sits roughly 8–10pp below where the non-GAAP measure is already heading, which is the right relationship. It is a demanding number but it is argued, and every alternative is priced in §1.4.

1.3 Exit multiple of 22.7x — growth-matched, named, and its limits stated

The requirement: the comparator set's growth must bracket the subject's growth at the exit year. TTAN compounding at ~20% for five years would be growing roughly 13–16% in year 5 on a normal SaaS decay path.

Comparator set — selected by business model before any multiple was inspected, from US-listed vertical / SMB SaaS, and filtered to those with a positive, non-outlier EV/EBIT:

Growth span: 11.0% – 20.2%. This brackets the 13–16% exit-year growth. The identification requirement is met. Median EV/EBIT 22.7x; mean 25.2x; range 14.2x – 34.1x, which is the sensitivity band in §1.4.

Excluded and why, stated so the selection is auditable: DOCN (70.2x), GWRE (266.4x), NCNO (570.0x) — EV/EBIT inconsistent with reported operating margin, indicating a depressed EBIT base; ZETA (867.1x) and GLBE (87.3x) same. APPF (26.3%) and PCOR (22.5%) sit above the exit-year growth band and were excluded to keep the bracket honest — including APPF at 36.8x would have raised the median.

The base exit multiple does not sit below every anchor. 22.7x is the median of the set and sits inside the 14.2–34.1x range, satisfying the rule that broke on NTRA.

Confidence: MEDIUM, and this is the weakest link in the memo. These EV/EBIT figures come from the same Tier-1 scanner that got TTAN's share count wrong by 2.7x. I verified TTAN's inputs against primary filings; I did not verify these seven. Their market caps are individually plausible against known scale, and the TTAN defect was specific to a recent IPO carrying a stale pre-IPO tag — but this is an unverified input and the sensitivity below is the correct way to consume it.

Implied compression, stated as a number: the required exit multiple of 3.81x EV/Sales (the sales-basis equivalent at 24.1% growth) against 6.96x today is a 45% de-rate. On EBIT: 48.8x on FY27 guided non-GAAP EBIT compressing to 22.7x is a 53% de-rate. Both are normal for a decelerating grower and neither is a heroic assumption.

1.4 Sensitivity — over the exit multiple, which is where the variance lives

Required 5-year revenue CAGR (%), WACC 10%:

Margin (demonstrated 24.1% − required), in percentage points — the number the strategy ranks on:

The single most important line in this memo: at the base exit multiple of 22.7x, the sign of the answer flips between a 16% and a 20% terminal margin. The verdict is a terminal-margin call, not a growth call or a multiple call. Anyone consuming this analysis should interrogate §1.2 and nothing else.

The screen's own 11.2% terminal margin sits below the worst row of this grid. On the corrected share count the screen's assumptions produce a required CAGR of 33.6% and a margin of −9.5pp — a FAIL.

WACC sensitivity (terminal margin 20%, exit 22.7x):

10% is the framework default, not a build. A 54%-realised-vol, zero-debt, GAAP-unprofitable software name plausibly carries an 11–12% cost of equity, at which the margin is +2.1 to +3.2pp. State the answer as "+2 to +5pp depending on discount rate," not as "+4.3pp."

1.5 Solving for the other unknowns, at demonstrated 24.1% growth

• Required terminal EBIT margin (exit 22.7x): 16.8% — versus 20% built. Cushion: 3.2pp of margin.
• Required exit EV/EBIT (terminal margin 20%): 19.0x — versus 22.7x used. Cushion: 3.7 turns, and 19.0x sits inside the observed comparator range (14.2–34.1x), so it is achievable on its own terms.
• Required exit EV/Sales: 3.81x. At a flat 6.0x exit EV/S (mild compression), the required CAGR falls to 13.3%; at 4.5x, 20.0%.

Every cross-solve lands in the same place: the price requires roughly what the business is currently doing, with a single-digit-percentage-point cushion that is sensitive to the terminal margin.

2. Own-multiple history — and why the 12-month anchor is UNIDENTIFIED

2.1 The window

ServiceTitan IPO'd 2024-12-12. The full price history is 405 trading sessions (1.61 years).

But the usable multiple history is shorter. A post-IPO TTM revenue figure did not exist in public until the FY2025 10-K was filed on 2025-04-02; before that the only public share count was the pre-IPO cover figure and the only TTM spanned a private-company period. The usable EV/Sales series is therefore 2025-04-02 → 2026-07-28: 331 sessions, 1.31 years.

Price statistics over the full 405 sessions: low $55.29, high $129.37, last $78.39 — the 28th percentile.

2.2 The declaration

The 12-month own-multiple anchor is UNIDENTIFIED.

Two independent grounds, either sufficient:

1. The window is 1.31 years — below the ~2-year minimum. 331 observations of a single continuous de-rating episode are one observation of a regime, not a distribution.
2. The sample spans a regime change, and it is a monotonic one. The series runs 14.67x → 5.03x → 6.96x. The "median" of 9.83x is not a level the stock mean-reverts to; it is the arithmetic midpoint of a one-directional collapse from post-IPO enthusiasm through lock-up expiry. Using it as an anchor would embed a +41% multiple expansion as the base case purely because the stock was once expensive. That is exactly the failure the rule exists to prevent.

No peer median is substituted. Per criteria.md and the brief, the anchor is declared unidentified rather than replaced. The comparator set in §1.3 is used for a five-year exit multiple, where growth-matching is defined and testable; it is not used for a 12-month target, where a peer median projected onto a different company over a mean-reversion horizon has no identification at all.

What would resolve it: roughly eight more months of trading takes the sample past two years and past the first full year of post-lock-up, index-inclusion-eligible float. Re-run then.

3. Near-term estimates — built off the demonstrated guidance-beat pattern

No consensus was available (Alpha Vantage quota exhausted; Consensus Criteria INDETERMINATE). The estimate base is therefore company guidance corrected by the company's own demonstrated beat behaviour, which is the most reliable thing in this entire memo.

3.1 Guidance vs actual — every quarter since the IPO

6 of 6. Mean +4.57%. Range +3.67% to +5.73% — a 2.1pp band across six quarters. This is one of the tightest and most systematic guidance patterns I have seen in this book. It is not a beat; it is a policy.

The same holds at the full-year level: FY26 was guided initially at $895–905m (mid $900m) in March 2025 and walked up to $910–920 → $935–940 → $951–953, printing $960.965m — +6.8% above the initial guide and +5.0% above the post-Q1 guide.

3.2 FY2027 house estimate

Current FY27 guide (post-Q1, 2026-06-04): $1,130–1,140m, mid $1,135m → +18.1% YoY.

Three independent methods land within $5m of each other. The house number is $55m (+4.8%) above guidance, and this is the memo's one genuine variant view. Its entire basis is a six-for-six sandbagging record. Non-GAAP operating income is sandbagged harder still — FY26 was guided at $48–53m initially and printed $94.058m — so the guided $142–147m for FY27 is very likely low, though I do not carry an explicit estimate because the FY26 beat magnitude (+86%) was inflated by first-year-public conservatism and will not repeat.

3.3 NTM revenue (Aug 2026 → Jul 2027)

Implied FY27 quarters on the house path: Q1 $268.8m (actual), Q2 $298.0m, Q3 $307.7m, Q4 $315.2m.

4. The 12-month target

4.1 The output

12-month target: UNIDENTIFIED as a point estimate.

Multiple-flat scenario: $90.41 – $95.96 (+15.3% to +22.4% to spot).

Per valuation.md step 3, a 12-month target must be anchored on the name's own trading range with the percentile stated. That anchor is unidentified (§2.2), and the brief is explicit that a peer median may not be substituted. Issuing a point target here would mean inventing the one parameter that determines it.

What I report instead is the estimate path with the unidentified parameter held explicitly free — the same discipline the implied-path test applies, applied to the same problem at a different horizon.

4.2 The conditional band

Assumptions: shares 100.2m in 12 months (+5.0% dilution, which is exactly what FY26 realised: 90.05m at 2025-01-31 → 94.60m at 2026-01-31); net cash $542m (current $421.5m plus ~$120m of free cash generation on the non-GAAP trajectory). Excludes the 6.48m Co-Founder RSUs, which only issue above $140 — see §4.4.

4.3 How to read this

The multiple-flat row is the load-bearing one, and it is not a forecast — it is arithmetic. If the multiple does not move, the 12-month return is approximately revenue growth minus dilution: +24% revenue, −5% dilution, plus a small net-cash build ≈ +15% to +22%.

That is the honest answer to "what is this likely to trade at," and it comes with an explicit statement of what it assumes: that the de-rating from 14.67x to 6.96x has stopped. It has been at or below 7.0x for part of the last quarter and the stock sits at the 31st percentile of its own (short, one-directional) range. Nothing in the data establishes that the de-rate is over. The −9.6% to −14.7% row is what a continued de-rate to the prior trough delivers even on 24% revenue growth, and it is not a remote scenario.

Expectation check. Item B16 recorded 16 of 16 house targets below spot. The multiple-flat scenario here is above spot by 15–22%, which is the normal and expected shape for a growing company, and the memo does not manufacture a below-spot answer by haircutting a multiple.

No external professional target is available for the sanity band; no Street coverage was retrievable.

4.4 The $140 discontinuity

If the stock reaches a $140 VWAP hurdle, 6,483,088 Co-Founder RSUs vest (Research §4.4) — +6.8% shares. Any scenario above $140 must be struck on ~101.9m shares plus the 5% organic dilution ≈ 106.7m, which is 6.4% dilutive to the price at that level. The 9.83x row above ($125–133) sits just below the trigger; the 14.67x row ($185–196) is well above it and should be read as ~$173–184 after the RSU dilution. This is disclosed here rather than buried because it is a mechanical, dated, disclosed dilution that most screens will miss for exactly the same reason they missed the share count.

5. Verdict summary

The memo issues no position verdict. The book decides.