What it is
A turnaround is a company the market has given up on — usually one posting GAAP (accounting) losses — that is nonetheless still generating real cash and has a credible path back to health. Think cyclical businesses at the bottom of their cycle, companies mid-restructuring, or names beaten down on a temporary shock.
The distinction that matters most: a GAAP loss is not the same as burning cash. Accounting earnings include non-cash charges — depreciation, write-downs, restructuring provisions — that can push reported profit negative while the business still collects more cash than it spends. A company losing money "on paper" but generating positive free cash flow is in a completely different league from one literally running its bank account down.
Turnaround = out of favor and unprofitable on paper, but still generating cash, with a path to recovery
Why it works
Turnarounds are where the market's overreaction is most extreme. When a company stumbles, sentiment doesn't just sour — it overshoots. The stock gets sold by funds that can't own "losers," abandoned by analysts, and priced as if the bad times are permanent.
But cyclical troughs are, by definition, temporary, and cash-generative businesses don't usually disappear — they recover. When they do, two things re-rate at once: earnings turn positive and the market re-applies a normal multiple to them. That double move is why a successful turnaround can return far more than a steady compounder over the same window. You are paid for buying when it's uncomfortable.
What "good" looks like
The line between a turnaround and a falling knife is cash. Signals that separate them:
- Positive free cash flow despite the GAAP loss — the single most important filter. The business funds itself.
- A cyclical or one-time cause, not structural decline — a commodity-price trough, a temporary demand air-pocket, a fixable operational stumble.
- Improving trajectory — margins, orders, or cash flow troughing and turning, not still deteriorating.
- A survivable balance sheet — manageable debt and maturities, so the company lives long enough to recover.
Common gotchas
- Falling knives. The mirror image of the opportunity. A company in structural decline — disrupted, obsolete, losing share permanently — keeps falling for good reason. "Cheap and hated" is necessary but nowhere near sufficient; the recovery thesis has to be real.
- Cash burn dressed as a loss. If the negative earnings come with negative free cash flow and a shrinking cash balance, this isn't a turnaround — it's a slow-motion solvency problem. The FCF filter exists precisely to exclude these.
- Leverage. Debt is what kills turnarounds before they turn. A highly levered company at a cyclical trough can be forced into a dilutive raise or worse before the recovery arrives.
- Catching it too early. Troughs can last longer than your patience. Sizing matters — these are higher-variance positions and should be weighted accordingly.
Turnarounds and the Bull Rankings score
The Bull Rankings model is a quality-growth score — it leans toward profitable, growing businesses trading at a fair price. By construction, that is not a turnaround screen: a company posting GAAP losses with a stalled top line scores poorly on the Quality and Growth pillars, however cheap it looks. So deep turnarounds generally won't rise to the top of the rankings, and that's intended — the score rewards businesses already working, not bets on a recovery that may or may not arrive.
That doesn't make the style invalid; it's a different, higher-variance game with its own discipline. If you pursue it, the one filter that separates a turnaround from a falling knife is the same one above: positive free cash flow despite the accounting loss. Genuine speculation — pre-profit, story-driven, cash-burning names — lives on the separate Moonshots shelf, which is explicitly outside the tracked model, and should be position-sized carefully.
Bottom line
Turnaround investing is buying what the market has wrongly left for dead — companies that look ugly on the income statement but still generate cash and have a real path back. The edge comes from the crowd's tendency to over-extrapolate bad times; the risk is mistaking terminal decline for a temporary trough. Insist on positive free cash flow, demand a real recovery catalyst, respect the balance sheet, and size for the volatility. Get those right and the survivors re-rate hard.
How turnarounds are handled here
There used to be a separate "turnaround" bucket in this model, selected alongside momentum and value buckets. It was retired in July 2026, and what replaced it is worth explaining because it reflects a change in thinking rather than a lost feature.
The buckets were replaced by a single GARP score applied uniformly to every name. The problem with a dedicated turnaround bucket is that it required the model to decide in advance which frame a company should be judged in — and that decision was doing more work than the analysis inside it. A company labeled a turnaround got graded on turnaround criteria and therefore looked like a reasonable turnaround. The label was producing the conclusion.
Where a recovering business shows up now
Loss-making companies aren't excluded. They get the same quality-growth score as every other company, measured with the same instruments. The only thing that changes is a label on the grade card: a company that is burning cash or has no positive P/E is marked speculative, and the card shows its price-to-sales ratio in place of a P/E.
There's no separate turnaround signal and no bonus for being one. That's deliberate — a label that decides in advance how a company should be judged is exactly what the retired bucket got wrong.
Several parts of the model give these names nothing, deliberately:
- The DCF produces no estimate at all. Discounting negative free cash flow yields a negative valuation, which tells you nothing you didn't know. No number is more honest than a confident one that means nothing.
- P/E is absent where there are no trailing earnings, so the value pillar leans on other reads.
- ROIC is usually negative, since operating profit is, so the quality pillar's heaviest read scores it at zero.
What the score can and cannot tell you about a recovery
The model is built to identify quality businesses at reasonable prices. A turnaround is, almost by definition, a business whose quality metrics currently look bad — that's what makes it a turnaround.
So a genuine recovery candidate will usually score poorly here right up until the recovery is visible in the filings, at which point it stops being a turnaround and starts being a normal company that the screen may well like. The model is a lagging indicator on exactly this kind of situation, and that's a structural limitation rather than a tuning problem.
If turnarounds are your strategy, the useful way to read this site is as a confirmation tool rather than a discovery one: when a name you've been watching starts clearing the quality and growth reads, the model is telling you the improvement has reached the financial statements. It will not tell you it's coming.