SPSC · SPS Commerce, Inc. · Stock research

Current Thesis

The leg on offer is a de-rating repair: a retail supply-chain network whose multiple broke, whose growth slowed to single digits, and whose profit line is now inflecting faster than its top line. On 2026-07-30 SPS reported Q2 revenue of $197.8M, +6% YoY from $187.4M, with adjusted EBITDA of $66.6M, +19% YoY from $56.1M, and non-GAAP diluted EPS of $1.27 against a $1.09 consensus. Management raised the FY26 adjusted EPS band to $4.84–$4.93 from $4.73–$4.76 while lowering the FY26 revenue band to $788.4M–$793.4M from $796M–$802M — the stated bridge being the 2026-06-30 divestiture of the 3P revenue-recovery business, which removes roughly $10.5M of second-half revenue at neutral impact to adjusted EBITDA.

What an investor is buying, then, is that the cash-earnings line is real, that the 1P supplier network is the durable asset, and that the July de-rating overshot. What sits against it is a specific, fresh bear thesis with a name attached: on 2026-07-21 Morgan Stanley's Chris Quintero cut SPSC to Underweight with a $57 target (from $70), arguing that agent-to-agent protocols let trading partners exchange documents directly and erode the case for an EDI intermediary, and that cross-sell from TIE Kinetix, Traverse Systems, SupplyPike and Carbon6 has not been demonstrated. That firm raised the target to $59 on 2026-07-31 and kept the Underweight.

Life-cycle: ACCELERATING, dated to 2026-07-30/31 — a Q2 EPS beat, three same-day target raises (Cantor to $70, Stifel to $70, Morgan Stanley to $59), and a session in which SPSC printed among the day's larger tech gainers. The qualifier matters: the acceleration is in price and in sell-side revision direction, not in revenue, which decelerated to 6%. The prior narrative — SPS as a premium-multiple serial compounder — is already dead; the 36.6% distance from the $115.82 52-week high is the receipt.

Bullish and bearish views on SPS Commerce, Inc.

The model's bull view on SPS Commerce, Inc. (SPSC), in brief: Profit inflects ahead of revenue. Q2 adjusted EBITDA $66.6M vs $56.1M a year earlier (+19%) on 6% revenue growth (2026-07-30 release). FY26 adjusted EBITDA guided to $264.6M–$269.1M, a ~34% margin at the midpoint. The revenue guide-down is mix, not demand — per the company. The… The bear view: Growth is at 6% and the near-term guide is below the street. Both cases follow in full.

Bull Case

• Profit inflects ahead of revenue. Q2 adjusted EBITDA $66.6M vs $56.1M a year earlier (+19%) on 6% revenue growth (2026-07-30 release). FY26 adjusted EBITDA guided to $264.6M–$269.1M, a ~34% margin at the midpoint.
• The revenue guide-down is mix, not demand — per the company. The 2026-06-30 sale of the 3P revenue-recovery business took roughly 7,300 customers and ~$10.5M of 2H26 revenue out of the model at neutral adjusted EBITDA impact (Q2 2026 call). Recurring-revenue customers stood at approximately 46,600 with average revenue per customer of $15,100 after the sale. Inference, flagged as such: consensus of $793.6M for FY26 revenue was set before the divestiture was in the numbers, so part of the "miss versus estimates" is a definitional gap rather than a demand gap.
• Capital returned into the drawdown. $51.2M of stock repurchased in Q2 2026 with $173.2M of cash and equivalents at 2026-06-30 (2026-07-30 release).
• Valuation has compressed to a level this name has not carried historically. Forward P/E of 14.4x and EV/EBITDA of 15.5x per stockanalysis.com as of 2026-08-08.
• Sell-side is split, not uniformly negative. Citigroup maintained Buy and raised its target to $82 on 2026-07-21 — the same week as the Morgan Stanley downgrade.

Bear Case

• Growth is at 6% and the near-term guide is below the street. Q3 revenue guided $196.3M–$198.3M against a $201.9M consensus; Q3 non-GAAP EPS $1.20–$1.23 against $1.24 (2026-07-30).
• GAAP earnings collapsed. Q2 GAAP net income $6.9M, $0.19 diluted EPS, against $1.27 non-GAAP. FY26 GAAP EPS guidance was cut to $2.24–$2.33 from $2.66–$2.69, versus a $2.59 consensus. A reader taking the 14.4x forward multiple at face value is paying for the adjusted line.
• The disintermediation argument is structural and slow to falsify. The A2A/agent-exchange thesis (Morgan Stanley, 2026-07-21) will not be settled by one quarter's ARR; it can suppress the multiple for several prints regardless of near-term execution.
• The M&A stack has not visibly converted. Four acquisitions — TIE Kinetix, Traverse Systems, SupplyPike, Carbon6 — with, per the same note, limited disclosed evidence of cross-sell. The June divestiture is a partial admission that at least one piece did not fit.
• Price trades above most post-print targets. With the last completed close at $73.38, spot sits above Morgan Stanley's $59, Cantor's $70 and Stifel's $70, and below only Citi's $82.

Setup & Price Structure

Last completed daily close $73.38 on 2026-08-07. 52-week high $115.82, leaving price 36.6% below it. Three-month return +29.9%. RSI(14) 60.8 — advancing, not stretched.

The structural event is the moving-average configuration: per stockanalysis.com on 2026-08-08, the 50-day sits at $60.83 and the 200-day at $70.04, placing the 2026-08-07 close above both. Reclaiming the 200-day after a 36.6% drawdown is the first evidence the downtrend has been interrupted rather than paused. The gap between spot and the 50-day is wide, which is what a near-vertical three-month recovery leaves behind; the better-defined structure is a retest of the reclaimed 200-day zone rather than the extension itself.

Crowding and positioning observables, stated without a verdict: short interest of 2.95M shares, 8.29% of float (stockanalysis.com, 2026-08-08) on 36.0M shares outstanding — elevated for a profitable, cash-generative software name, and a mechanical contributor to the post-print move. Five-year beta 0.49, so the move is idiosyncratic rather than market-carried. No earnings date inside the next 30 days, which removes the usual pre-print bid. Retail-sentiment coverage of this name is thin; the attention arriving is sell-side, and three of the four post-print target revisions land below the last close.

Catalyst Calendar (next 30 days)

• No company-dated event falls inside 2026-08-08 → 2026-09-07. Q2 was reported 2026-07-30; the Q3 release date had not been announced as of 2026-08-08. The window is empty and the stock is left to trade on drift and positioning.
• ~2026-10-22 (est.) — Q3 FY26 print. Outside the window, but it is the first quarter that fully reflects the 3P divestiture and the first test of whether recurring revenue holds ~6% ex-disposal.
• ~2026-10-09 (est.) — company announcement of the Q3 earnings date. SPS has historically posted the date roughly two weeks ahead of the release; this is the first scheduling marker to watch.

What Would Change Our Mind

The structure that carries this read is the reclaimed 200-day near $70.04 and the June–July shelf beneath it. Losing that shelf would mean the three-month advance was a relief bounce inside a downtrend and that the multiple compression has further to run: a weekly close below $70 ends the repair read, with the $57–$59 Morgan Stanley target zone as the next reference.

Three non-price conditions would independently break it: Q3 revenue printing at or below the $196.3M low end of guidance, which would say the deceleration is not divestiture arithmetic; a further cut to the FY26 revenue band in October after this one; or a named top-tier retail trading partner publicly routing document exchange through an agent protocol rather than an EDI network, which would move the Morgan Stanley thesis from argument to evidence. Conversely, if the theme flips to SATURATED — mainstream "cheap software bottom" coverage, a cluster of upgrades to Buy, price extending well past $82 without a fundamental revision — the reward for the same risk has left.

Correlation Notes

• Trades with the SMID-cap application-software de-rating cohort rather than with AI infrastructure; the 0.49 five-year beta means index direction explains little of the variance.
• Second-order exposure to US discretionary retail and consumer-goods supplier health: recurring revenue is a function of trading-partner count and per-partner document volume, so a retail-inventory downcycle transmits with a lag rather than immediately.
• Directly exposed to the agentic-protocol narrative complex. Headlines advancing agent-to-agent commerce standards are a live re-pricing risk for this name specifically, independent of its own results — which is what makes the Morgan Stanley note of 2026-07-21 a recurring, not a one-time, overhang.
• The GAAP/non-GAAP spread ($0.19 versus $1.27 in Q2) means screens built on GAAP earnings and screens built on adjusted earnings will place this name in opposite buckets, which is one mechanical source of the split sell-side.