When earnings improve in a company that was under pressure, the room changes. The CFO presents a recovered margin. The CEO talks about momentum. The sponsor exhales. Within a few weeks, attention starts to drift back to the strategic agenda — the M&A idea that was paused, the geography expansion, the platform investment.
This is, in our experience, the most dangerous point in any restructuring. Not because the recovery is fake, but because it is being read as evidence that control has returned. EBITDA does not prove control. It records the consequence of actions taken — repeatable or not.
1. The improvement may not be repeatable
Most early margin recoveries are produced by actions that have a ceiling. A workforce reduction. A discretionary spending freeze. A round of supplier renegotiation. Inventory liquidation. Deferred capex. A pricing pass that the market accepted once.
Each of these moves real money to the bottom line. None of them produces a structurally different company. After two or three quarters, the same drivers that produced the original underperformance are still there — slower, perhaps, but unchanged. The improvement looks like recovery; it is actually a one-off rebalancing.
The timing of the misreading is itself a pattern. The good quarter is presented to the Board roughly at the moment the one-off measures have exhausted their effect. The next quarter is already trending back toward the prior baseline, but that information will not reach the Board for another two months. By then, attention has shifted, the strategic agenda has reopened, and the conversation about the original problem has lost its weight.
2. EBITDA does not say whether cash will follow
In stressed companies, EBITDA and cash diverge for structural reasons. Working capital absorbs receivables that have been promised but not collected. Inventory builds because nobody wants to call a price reset. Accruals hide unresolved decisions. By the time the operating accounts close, the EBITDA looks recovered and the cash position has not moved — or has continued to deteriorate.
If margin improvement does not convert reliably to cash, equity value is fragile, regardless of how the income statement reads. The lender, the buyer, and the auditor all know this. The room often does not, because the operational reporting that would surface the divergence has not been rebuilt.
The aggregate picture supports the warning. The Hackett Group's 2025 European survey records cash conversion deteriorating across the continent even as many companies defend or recover margin: days-sales-outstanding up, days-inventory-outstanding at decade highs, total excess working capital in the order of €1.4 trillion. Margin recovered without working capital recovered is, in the lender's reading, a story still missing its second half.
3. EBITDA does not prove the company can predict itself
Control is not a level. It is a condition: the company knows what is going to happen before it happens, and it can correct course before the variance compounds. A single recovered quarter is not evidence of this. It is evidence that one quarter went better than the previous one.
The relevant signal is not the absence of surprises — there will always be surprises. It is how the system reads them and reacts. A company in control sees the variance early, traces it to a decision or a driver, and adjusts within the cycle. A company that has only recovered EBITDA experiences the same variance late, debates whose number is correct, and reacts in the next quarter, after the gap has already compounded.
Operationally, the difference is visible in places that do not appear on the income statement. How fast does the rolling forecast incorporate a new piece of information from the field? Does the management team agree on the operating reality before they get into the meeting, or do the disagreements surface in front of the Board? When a customer delays an order or a supplier shifts terms, does that information move through the company in days or in weeks? Is there a single owner for each cash-relevant decision, or does ownership shift depending on who is in the room?
Companies that pass these tests stay in control through the next shock. Companies that don't, lose the recovery within two quarters.
What control actually looks like
Restoring control is a different exercise from recovering EBITDA. It requires earnings quality (one-offs separated from structural moves), cash conversion that is reliable enough to plan against, a cost architecture aligned to demand reality, and a decision cadence that surfaces variances early enough to act.
When these are in place, the next margin improvement is durable, because the system that produced it has the discipline to defend it. When they are not, every recovery is provisional — vulnerable to the next external shock, the next leadership change, or the next moment the Board stops asking.
EBITDA is a metric. Control is a condition. The mistake is to read the first as evidence of the second — and to release the pressure on the management team exactly at the moment when the system is still fragile and most needs it.
There is also a longer-cycle reason to keep the pressure on. Earnings quality is not something that can be assembled in a data room before an exit. McKinsey's 2026 review of private equity highlights how, with hold periods extending and over half of the global buyout-backed inventory now held more than four years, every quarter of recovery is already part of an eventual diligence file. A margin restored without the underlying control is, by the time the buyer arrives, easy to read.
---
References
- 2025 European Working Capital Survey: Cash Cycle Deterioration, The Hackett Group, November 2025.
- Are Legacy Metrics Derailing Your Transformation?, Harvard Business Review, February 2026.
- Global Private Equity Report 2026: Clearer view, tougher terrain, McKinsey & Company, February 2026.