The 0.87x P/S and 3.02x P/FCF appear optically inexpensive, but the comparison to ANIP's 9.67x EV/EBITDA and Halozyme's 19.28x is not like-for-like, since those peers sustain positive operating margins while Akebia's swung back to -12.9% in 2026Q2. The 9.55x EV/EBITDA is the more defensible anchor, yet with EBITDA that has flipped between positive and negative across the last ten quarters, investors should treat it as a point estimate rather than a normalized figure. The 6.03x price-to-book may also overstate real asset backing, given that $59.0M of the $349.7M asset base is goodwill against $25.0M of equity.

The gap between a 78.8% gross margin and a -18.1% net margin implies roughly 31 points of operating and below-the-line cost against a $49.1M revenue base, which suggests fixed SG&A and R&D absorption is the binding constraint rather than unit economics. 2025Q1's 23.6% operating margin and 10.7% net margin appear to have been a volume-driven peak rather than a sustainable run-rate, since each subsequent quarter reversed toward losses. Adjusted for the 2025Q4 R&D spike, true earning power still looks closer to breakeven than to the peak print, which materially narrows the margin of safety in the current valuation.

The one observed ROIC readings — 38.2% in 2025Q1 against -39.7% in 2024Q4 — are too far apart to establish a trend, and with ROA pinned between -2.5% and -3.3% over the last three quarters, the asset base is generating essentially no return. Because equity has moved from negative $49.2M to $25.0M, ROE is being driven as much by the denominator's recovery as by any earnings improvement, which means the -34.0% print may overstate the severity of capital destruction relative to the modest absolute loss. On the available evidence, the company appears to be decaying rather than compounding returns, and investors should monitor whether ROIC becomes consistently reportable before drawing firmer conclusions.

DSO has climbed from 66 days in 2024Q4 to 107 days while DIO sits at 114 days versus a 691-day (distorted) 2024Q1 figure and 168 days in 2025Q3, suggesting customer collections are lengthening faster than the company is stretching its own suppliers. Asset turnover of 0.14x, down from 0.22x in 2024Q4, implies each dollar of assets now supports only fourteen cents of annual revenue, which is the mechanical reason ROA remains near zero. The 143-day cycle sits below the 338-day 2024Q1 extreme but well above the 84-day 2024Q4 trough, so working capital appears to be a recurring source of both cash inflows and outflows rather than a stable efficiency lever.

The leverage jump appears driven largely by a shrinking equity denominator rather than a step-change in borrowing, which means the ratio's trajectory will track loss persistence as much as any new financing decision. Coverage has been negative in seven of the last ten quarters, the exceptions being 2025Q1-Q3 at 1.04-1.79, so the brief 2025 profitability window was the only period in which debt service looked comfortable. With D/EBITDA unreportable in recent quarters because EBITDA is near or below zero, refinancing risk is difficult to size, and investors should monitor whether coverage can re-approach 1.0x before the equity cushion erodes further.