Under Armour: The Last Layer

Under Armour has stripped away the athletes, the apps, the SKUs, and the growth story. What's left is a $5 stock, an 84% one-star rate, and a value investor betting $200 million that the base layer still fits.

On May 12, 2026, Kevin Plank stood on Under Armour's earnings call and declared the turnaround's "most disruptive phase" complete. He listed what had been cut: 25% of all product styles. The Rialto distribution center. The chief product officer. The chief financial officer. The chief people officer. The Curry Brand — an entire basketball franchise built around one of the greatest shooters in history, unwound in six months, the last Curry 13 colorways trickling out through October.

Plank called what remained "the foundation." A company stripped to its core, entering what he described as an "execution and simplification" phase. Revenue had shrunk from $5.7 billion to $5.0 billion over two fiscal years by design. The plan was to emerge leaner, more premium, more focused.

Eight days later, a customer in Liverpool posted a one-star review on Trustpilot. She had returned $270 in merchandise via DHL six weeks earlier. Under Armour told her the parcel arrived empty. She had the tracking confirmation showing otherwise. After four calls, she filed a chargeback.

Her review was the 1,130th one-star review out of 1,345 total on Under Armour's US Trustpilot profile. That is an 84% one-star rate. On SiteJabber, an independent platform, the one-star rate is also 84%. We ran a two-proportion Z-test: Z = 0.12, p = 0.904. The rates are statistically indistinguishable. This is not a sampling accident. This is a structural customer experience failure confirmed across independent data sources.

Plank's turnaround has a blueprint. The blueprint looks credible on paper — SKU reduction, premium positioning, a new "Sports House" retail concept at Tysons Corner that carries only 96 styles. But blueprints are drawn in offices. Customers live in basements. And Under Armour's basement — the returns infrastructure, the e-commerce fulfillment, the post-purchase experience — is flooded.

The question is whether the blueprint can outpace the flood.

See the Investment Council's verdict on $UAA → | All current verdicts: The Verdict Board

The shape of the decline is older than Plank's return.

Under Armour went public at a split-adjusted $3 in 2005. By December 2015, the stock hit $52 — the original performance brand of the post-Nike generation, endorsed by Steph Curry, Tom Brady, and Jordan Spieth, with a connected fitness empire built on the $150 million MapMyFitness acquisition. Revenue had grown every quarter for nearly a decade.

Then the cracks opened. In January 2017, Plank called President Trump's business-friendliness "a real asset." Curry publicly distanced himself. Boycott hashtags trended. The culture war stuck to the brand like sweat to cotton — exactly what Under Armour promised its fabric wouldn't do.

The business metrics followed. North American revenue, 77% of total sales, began declining in 2017 and never fully recovered. By 2020, Plank had stepped down as CEO. His replacement, Patrik Frisk, lasted two years. Stephanie Linnartz, recruited from Marriott, lasted eighteen months. When Plank returned in April 2024, the stock was at $7 and the company had cycled through three CEOs in four years.

What Plank inherited was a brand in category limbo. Under Armour had ceded the running market to Hoka (On Holdings' revenue grew 30% in FY2025 on $3.4 billion). It had ceded premium athleisure to Lululemon ($11.1 billion, 56.6% gross margins). It had ceded basketball credibility to Nike's Jordan brand. And it had ceded its own digital moat by selling MapMyFitness to Outside Interactive in August 2024 for undisclosed terms — a platform it had paid $150 million for in 2013, with 80 million registered users, divested as a non-core asset.

The stock today: $5.58. Down 89% from its all-time high. Down 31.5% from its 52-week high of $8.15. Market cap: $2.38 billion — less than half of what On Holdings trades at on two-thirds the revenue.

*FY2027 guidance includes $70M in IEEPA tariff refund benefits. Excluding the refund, adj. OI would be $70-90M.

The gross margin trajectory is the headline risk. After peaking at 48.2% in Q1 FY2026, margins collapsed sequentially each quarter — 47.3%, 44.4%, 42.0% — driven by escalating tariffs, higher product costs, and unfavorable regional mix. Full-year FY2026 gross margin of 45.5% was 240 basis points below FY2025.

Management frames FY2027 as a "stabilization year" with FY2028 as the growth inflection. But the FY2027 adjusted operating income guidance of $140-160 million includes that $70 million tariff refund — a one-time benefit that represents 44-50% of the lower bound. Strip it out, and the operating income is roughly what Under Armour earned in 2019, when revenue was $1.2 billion higher.

The balance sheet is manageable but weakened. Cash fell from $501 million to $309 million year-over-year. The company refinanced $600 million of 3.25% notes maturing in 2026 by issuing $400 million of 7.25% notes due 2030 — more than doubling the coupon rate on that tranche. Moody's downgraded the credit rating to B1, projecting debt/EBITDA at 5.5x.

By segment: North America ($2.9 billion, -8%) continues to contract while international ($2.1 billion, +4%) grows, led by EMEA and Latin America. Footwear ($1.1 billion, -11%) is the weakest category. The DTC channel is shifting from e-commerce (-7%) toward owned stores (+1%).

Total quantitative sample: ~4,526 reviews + 120 complaints across 7 independent platforms. The Trustpilot UK sample (~1,320 reviews) provides geographic independence from the US data.

The surface comparison is deceptive. Both Under Armour and Nike sit at 1.6 out of 5.0 stars on Trustpilot. Both are punished by e-commerce customers. The temptation is to call this an industry-wide problem and move on.

But the distribution tells a different story. Under Armour's one-star rate is 84%. Nike's, estimated from its 12,000-review Trustpilot profile, is approximately 79%. We ran a two-proportion Z-test.

H₀: Under Armour's 1-star proportion = Nike's 1-star proportion.

Result: Z = 4.26, p < 0.001. The difference is statistically significant.

Under Armour's 1-star rate is 4.9 percentage points higher than Nike's (95% CI: 2.8 to 7.0 pp). On the same platform, at the same headline rating, Under Armour has a significantly more extreme complaint concentration than its closest peer.

The dominant complaint pattern is returns/refunds dysfunction. Not product quality, not shipping speed — the post-purchase infrastructure. DHL loses the return. Under Armour claims the package arrived empty. Refunds stall for 30+ days. Customers file chargebacks. BBB complaints show the same pattern: 57% are product issues, but the resolution path (escalation through 6+ customer service representatives, apology gift cards instead of refunds) reveals a systemic operational failure.

If the turnaround is working, complaints should be declining. The BBB data provides a testable time series: 120 complaints over 36 months, with 26 in the last 12 months versus 94 in the preceding 24 months.

H₀: Complaints arise from a homogeneous Poisson process (constant rate).

Result: p = 0.007. The rate change is statistically significant.

BBB complaints have declined from 3.9 per month (months 1-24) to 2.2 per month (months 25-36) — a 44.7% drop. This is the single most encouraging signal in the data. The operational infrastructure that generates complaints appears to be improving, even as the review platforms still reflect accumulated historical damage.

The interpretation requires care. Trustpilot and SiteJabber ratings are cumulative — they reflect years of reviews, including periods before the turnaround began. The BBB complaint rate is a flow metric, not a stock metric. A declining complaint rate does not mean the brand is healthy; it means the rate of new damage is falling. The 84% one-star cumulative rate will take years to dilute even if every new review is five stars.

The financials do not capture what happened to Under Armour's moat.

In 2015, Under Armour had three competitive barriers: (1) Stephen Curry, the most marketable athlete in basketball, whose Curry 4 outsold every non-Jordan signature shoe; (2) MapMyFitness, 80 million registered users feeding a connected-fitness data flywheel; and (3) a "performance first" brand identity that justified premium pricing in compression, base layers, and training gear.

All three are gone. Curry signed a 10-year deal with Li-Ning after growing "frustrated" with what he perceived as underinvestment in his line. MapMyFitness was sold to Outside Interactive. And the "performance first" positioning has been muddled by a pivot toward lifestyle — a Parker McCollum country-music ambassadorship, a "Bouncy Tee" campaign with rapper Gunna — that reads more like a Hail Mary than a brand strategy.

What remains is recognition without preference. Under Armour's Net Promoter Score of 30 ranks seventh among major sportswear brands — behind Nike, Adidas, New Balance, Puma, Reebok, and ASICS. The brand is universally known and increasingly unchosen.

The Glassdoor data quantifies the internal disconnect. Employees rate Under Armour 3.7 out of 5.0 — reasonable. But customers rate it 1.5 out of 5.0 on Trustpilot. That gap of 2.2 points is massive: Welch's t = 55.2, p < 0.001, Cohen's d = 1.81. Employees see a company in recovery. Customers see a company that can't process a return.

Recovering:

NOT recovering:

Unknown:

1. Trustpilot and SiteJabber skew negative by design. Satisfied customers rarely post reviews on complaint-oriented platforms. The 84% one-star rate reflects the composition of who reviews, not the composition of all customers. However, the UA-vs-Nike comparison controls for this bias (same platform, same category), and UA is still significantly worse.

2. The BBB complaint decline could reflect reporting fatigue rather than operational improvement. If customers learn that BBB complaints lead to gift-card apologies rather than refunds, they may stop filing — reducing the rate without fixing the problem.

3. Glassdoor's 3.7 rating includes historical reviews. The business outlook metric (43% positive) is more diagnostic of current employee sentiment than the headline rating.

4. Nike comparison is imperfect. Nike has 8.9x more Trustpilot reviews (12,000 vs 1,345), and its scale means a different operational profile. The two-proportion Z-test controls for sample size but not for business-model differences.

5. Financial projections depend on tariff assumptions. The $70M refund in FY2027 guidance represents 44-50% of the low end of adjusted operating income. Any change in trade policy could materially alter the outlook.

Under Armour is a bet on whether Kevin Plank's "shrink to grow" strategy can generate a premium brand from a discount stock. The bull case has a deep-value floor in Fairfax Financial, improving operational metrics (declining complaints, growing international), and a management team that has executed the painful cuts on schedule. The bear case has a third consecutive year of revenue decline, a vanishing moat, tariff dependency, and the worst customer sentiment in its competitive set.

The probability-weighted expected value of $4.90 sits below the current price of $5.58. The stock is not cheap enough to compensate for the execution risk, unless you weight the Fairfax takeout scenario higher than 10% or believe the international growth trajectory can offset the North American decline faster than the consensus expects.

Now ($5.58): Under Armour trades at roughly 47-70x FY2027 guided EPS of $0.08-$0.12. That multiple only works if FY2027 is a trough. If Plank's timeline holds and FY2028 delivers $0.25-$0.35 in EPS, the stock is trading at 16-22x forward earnings two years out — reasonable for a recovering consumer brand. If the timeline slips, the multiple expands into a territory that has historically killed turnaround stocks.

Next catalyst: Q1 FY2027 earnings (early August 2026). Management guided Q1 as the weakest quarter — revenue down 2-3%, EPS of $0.00-$0.02. But Q1 also carries the heaviest tariff refund benefit (gross margin guided up 610-630 basis points YoY). If the refund lands and margins expand as guided, the stock has room to recover from the 17% post-earnings selloff. If the refund is delayed or reduced, the FY2027 guide collapses.

Decider date: March 2027 — full-year FY2027 results. This is when "stabilization year" gets tested. Does North American revenue stabilize? Does the Sports House concept generate measurable same-store sales lift? Does the international growth trajectory accelerate enough to offset domestic contraction? FY2027 is the year Plank said the foundation would be complete. If the foundation is still cracking, the FY2028 growth inflection becomes a mirage.

When Q1 FY2027 earnings land in early August, we'll publish an updated analysis with:

1. The tariff refund check — did the $70M IEEPA refund benefit show up in Q1 gross margin? If Q1 gross margin expands the guided 610-630 bps, the FY2027 guide is credible. If not, the full-year numbers need to be re-baselined.

2. North America trajectory — management guided Q1 NA revenue down 2-3%. The range matters: -2% suggests stabilization; -5% suggests the decline is accelerating despite SKU reduction.

3. Sports House read-through — the Tysons Corner store opened March 26. By August, Under Armour should have initial traffic, conversion, and basket-size data. Whether management shares it, and what it shows, will signal whether the format can scale.

4. Fairfax accumulation update — Prem Watsa's buying pace is a real-time read on deep-value conviction. Any acceleration (or pause) in purchases will be disclosed in SEC filings by August.

The blueprint says Under Armour is worth $7. The basement says $3. The question is whether Plank can drain the basement before the blueprint gets wet.

See the Investment Council's verdict on $UAA → | All current verdicts: The Verdict Board

Turnaround Radar is a research publication. Nothing here is investment advice. We don't know your financial situation, risk tolerance, or investment objectives. Do your own work.