The same hospital group, two buyouts, two outcomes

Australia's second-largest private hospital operator was bought by private equity twice, nine years apart. The first owners grew it by a third, floated it, and were reported to have doubled their money. The second bought it back off the market at a higher price, and six years later the lenders had it.

We tested both deals with our own deal assessment, in the same exploratory, AI-assisted retrospective test. The first came out as a case to keep going with. The second came out as a case to stop. The difference was on the page.

Healthscope is a useful case because it is the same business both times. Same hospitals. Same funding model. Same regulators. The public information was the same kind too: a listed company with three years of results, a bid contest, and a booklet sent to shareholders.

Four things changed between 2010 and 2019. The track record. The price. Who owned the hospital buildings. And the plan.

All four were public.

2010: a growing business at ten times earnings

In October 2010 TPG and Carlyle bought Healthscope after a contested auction. They paid a premium of nearly forty per cent, and about ten times the last year's earnings.

Earnings had grown at close to twenty per cent a year for five years. The growth came from more admissions, more people holding private health insurance, and beds added inside the company's existing hospitals. Thirty of its forty-four hospitals were owned outright.

The new owners raised public debt within weeks, so the borrowing was stated in dollars. It came to just under five times earnings, against a limit of five. The prospectus showed how far earnings could fall before the debt was in trouble. The headroom was tight, and it was visible.

An executive chairman who had run two hospital groups before was appointed within weeks. Four financial bidders and a trade buyer had just competed for the company. That answered the question of who might buy it later.

One part went badly. The pathology business, which the buyers had not wanted, depended on government fees. The fees were cut in the year of the deal. Its margins halved and its goodwill was written off.

The hospitals performed. Earnings grew by about a third in four years. The company returned to the stock market in 2014 at the top of its price range.

The owners had sold out by the end of 2015. Their actual return is not public. Reuters reported at the float that they stood to double their money.

2019: a shrinking business at fourteen and a half times

In June 2019 Brookfield bought the same company back off the stock market. It paid fourteen and a half times the last year's earnings, again at a premium of forty per cent.

This time earnings had fallen in the last year. Hospital margins had fallen three years running. Guidance had been cut three times, and two hospitals had closed. The company's own results slides said wages were rising faster than insurers would pay. They said fewer people were holding private health insurance.

The bid announcement promised a return to growth, with earnings up by more than ten per cent. Three months later, before the deal completed, that guidance was cut.

The price was paid for partly by selling the buildings. Twenty-two of the forty-three hospitals were sold to two landlords on the day the deal signed. They were leased back for twenty years. The starting rent was about a hundred and twenty million dollars a year, rising every year.

An independent expert reported to shareholders on the deal. The expert named the sites as the company's advantage, "extremely difficult to replicate". The same report warned that renting them back would leave higher operating leverage and more volatile earnings. The booklet admitted the debt was higher than the company had carried before, and "in line with recent leveraged buyout transactions in the healthcare sector".

Six years on, as at September 2026, this is the record. A pandemic halved surgical activity. The company had three chief executives.

The goodwill was written down to nil, for "an unsustainable cost structure". Contracts with insurers covering millions of members were ended and then restored. A public hospital the company ran was handed back to the state.

Then rent was missed. Receivers were appointed in May 2025, over more than A$1.6 billion of debt. Lenders were estimated to recover about forty cents in the dollar.

The landlords found new operators for their hospitals within months. The advantage survived. The company that had sold it did not.

What our assessment says about each deal

Our fund assessment asks nineteen questions about a deal, in five areas. Each answer sits on a five-step scale: red flag, doubtful, unproven, supported, proven.

We answered the questions twice for each deal.

The first time, we used only what a sceptical reader could have found on the day of the deal. The company's results. The bid documents. The shareholder booklet. No hindsight.

Sixteen of the nineteen questions could be answered this way. The other three need the buyer's private plan, so we left them blank and did not mark them down.

The second time, we used everything known today.

The 2010 deal, on the day. A sound case with gaps. Seven answers were "supported", eight were "unproven", and one was "doubtful".

None was a red flag. None of the warning questions scored low. Our verdict would have been: keep going.

The 2019 deal, on the day. Nine of the sixteen answers were "doubtful". Three were "supported", and those were about what the business is, where it operates and what its advantage is. Most of the warning questions scored low. Our verdict would have been: stop.

With hindsight, the gap widens. The 2010 deal moves up. Most answers become "supported" and two become "proven". The 2019 deal moves down. Ten of its sixteen answers become red flags.

So the assessment told the two deals apart without the outcome. It separated them on the day, from public documents. Hindsight only confirmed it.

What changed between the two deals

In the terms of our piece on why deals fail, the 2019 deal had a cause and three amplifiers. The cause was that someone else set the price. Insurers and governments decided what the hospitals were paid, while wages rose faster. The amplifiers were a contest price, debt sized to the plan, and buildings sold and rented back.

What a buyer can take from it

Believe the track record, not the announcement. A company that has cut its guidance three times is showing you what it earns. Price it on that. Before paying for a recovery, ask what has changed to cause one.

Do not sell the thing you are buying. The expert said the sites were the advantage. The funding plan sold them. Check that the way you pay for a deal does not remove the reason for doing it.

Treat rent as debt. A twenty-year lease with rising rent has to be paid in bad years, exactly like interest. Add it to the borrowings, then test whether the business survives a bad year. Here the low case still showed growth, so nobody had tested it.

Read the documents you already have. None of this is hindsight. Every fact above was in the company's results, the bid documents or the shareholder booklet before the money moved. No data room was needed.

Our assessment takes about ten minutes. Answered honestly, it would have said keep going in 2010 and stop in 2019. A committee that had seen "stop" would have had a different conversation.

• The deal checks: buying as a fund, buying as a strategic acquirer, and selling
• Why do private equity deals fail?
• How a Stages of Excellence check works