Amber International (AMBR) Q2 2026 earnings review
AI Pivot Yields Immediate Margin Reversal, But Visibility Goes Dark
Amber International's aggressive pivot from a digital asset platform to an 'Agentic AI' company is showing early financial validation. Q2 revenue reversed its downward trend, jumping 38.8% sequentially to $13.9M. The true story, however, is the margin profile: the launch of Agentic Market Making (A-MM) drove gross margins to a record 79.5%, pulling Adjusted EBITDA back into positive territory ($1.9M). However, the legacy execution and payment segments continue to collapse under macro pressures, and management completely withdrew forward guidance. The company is fundamentally changing its identity, leaving investors with a high-margin proof of concept but zero future visibility.
🐂 Bull Case
The initial launch of Agentic Market Making (A-MM) immediately contributed $3.5M in high-margin revenue, proving the commercial viability of the AI pivot and instantly expanding gross margins by nearly 1200 basis points sequentially.
Despite a 39% QoQ revenue surge, total operating expenses remained perfectly flat at $10.0M. Management's use of their own AI agent (MIA) is successfully decoupling revenue growth from headcount expansion.
🐻 Bear Case
Management withdrew all financial guidance. For a company transitioning business models, asking investors to fly blind creates massive uncertainty and undercuts the narrative of a seamless AI transition.
Execution Solutions revenue crashed 63% YoY to just $0.7M, and Payment Solutions fell 40%. The macro environment for digital assets remains soft, meaning the new AI segments must carry the entire growth burden.
⚖️ Verdict: ⚪
Neutral. The immediate return to profitability and gross margin expansion are highly impressive. However, the complete withdrawal of financial guidance and the collapse of legacy crypto segments make it impossible to model the second half of the year with confidence.
Key Themes
Agentic Market Making (A-MM) is the New Profit Engine
The successful rollout of A-MM is the primary driver of this quarter's turnaround. Generating $3.5M in its initial recognition, it immediately altered the company's margin profile. Because A-MM operates with software-like incremental costs rather than human-capital intensive wealth management costs, it drove gross margins to 79.5%. This validates management's thesis that 'Agentic Finance' is significantly more capital efficient than their legacy model.
Operating Leverage via Internal AI Implementation
Amber is effectively drinking its own champagne. By deploying MIA (their marketing and growth agent) internally, they held total operating expenses perfectly flat at $10.0M despite a $3.9M sequential jump in revenue. General and administrative expenses barely budged from $6.15M to $6.30M. This demonstrates accelerating operating leverage and proves their AI agents can physically replace traditional SG&A cost scaling.
Wealth Management Rebound
While overshadowed by the AI narrative, the legacy Wealth Management Solutions segment showed a solid reversing trend, growing 24.8% QoQ to $5.3M (up from $4.2M in Q1). Management attributed this to stronger demand across diversified investment products, establishing a stable floor for the core business while the AI segments ramp up.
Complete Withdrawal of Forward Guidance
In a highly concerning move that contradicts the glowing AI narrative, management completely withdrew all previously issued financial guidance. They stated that old metrics are 'no longer an appropriate measure' due to the transition. If A-MM and Ambre are scaling as predictably as claimed, management should be able to forecast them. Refusing to issue guidance signals internal uncertainty about the actual run-rate of these new revenue streams.
Macro Pressures Crushing Execution and Payment Solutions
The legacy transaction-based businesses are decelerating rapidly. Execution Solutions plummeted 63.3% YoY to just $0.74M, and Payment Solutions dropped 40.0% YoY to $0.52M. Management directly blamed broader macroeconomic conditions and risk-off positioning in the digital asset market. If this macro weakness bleeds into the Wealth Management segment, the AI growth could be entirely offset by legacy decay.
Aggressive Share Repurchases Provide a Price Floor
Management is putting its balance sheet to work. They executed $5.8M in buybacks (2.64M ADSs) out of their $50M authorization. With $44.2M remaining and a cash balance of $34.2M, the company has the liquidity to aggressively defend its stock price during this transition period, signaling strong internal conviction.
Other KPIs
Accelerating. An incredible improvement from 67.7% in Q1 2026. This was achieved via mix shift, heavily skewed by the initial high-margin revenue recognition of the A-MM business.
Stable. Up slightly from $33.9M at the end of FY25. The balance sheet remains highly liquid and completely unencumbered by traditional debt, providing ample runway for the AI transition and continued share repurchases.
Guidance
Reversing previous transparency. Management formally withdrew all financial guidance, including the previously communicated Amber Premium revenue outlook, citing the inability to forecast the new agentic AI business initiatives accurately. They intend to provide updated guidance only 'once sufficient operating history' is established.
Key Questions
A-MM Revenue Sustainability
The $3.5M recognized for A-MM this quarter drastically improved margins. Was this entirely recurring flow revenue, or did it include any one-time setup/licensing fees? What is the baseline run-rate we should expect going into Q3?
Guidance Reinstatement Timeline
You withdrew guidance to evaluate the new business initiatives. What specific internal KPIs or operational milestones do you need to hit before you feel comfortable reinstating forward revenue and EBITDA guidance for the street?
Legacy Business Floor
Execution and Payment solutions are down 63% and 40% YoY, respectively, due to macro headwinds. At what point do you consider sunsetting or spinning off these legacy crypto functions to focus purely on the AI Agent architecture?