Where buyout value really comes from
Research: 33 leveraged buyouts, rebuilt from their filings
Richard Parry
Every buyout is underwritten on a story about where the return will come from: growth, margin, cash, a better exit multiple, or leverage doing the heavy lifting. Once the deal closes, the story is rarely checked against what actually happened.
So we checked. atlanticma.ai's LBO corpus study rebuilt 33 public leveraged buyouts, closed between 2003 and 2016, from the buyers' and targets' own SEC and Companies House filings. They include HCA, Hilton, Dollar General, First Data, Burger King, Nielsen, TransUnion and Debenhams.
What the study did
Each deal's filed terms and filed operating results went into the LBO model, and the model's output was compared with what the filings say happened: the debt and cash at exit, and the sponsor's return.
The standard is 1%. Net debt at exit lands within 1% of the filed figure on 31 of 33 deals, and the sponsor's return on 30 of 32. The two deals outside that standard, Debenhams and Hilton, are published, not dropped. Where the model and a filing differ, the difference is declared with its cause rather than adjusted away: 42 of them across the corpus, such as figures no filing ever printed.
Each deal's gain is then split into a value bridge. The operational legs are what the business did: sales, margin, cash generation, acquisitions. The financial legs are what the deal structure did: the change in the multiple, recapitalisations, disposals and fees.
What it found
- Median sponsor return, as filed, across 32 deals
- 2.44x
- Deals where EBITDA growth was the largest source of value
- 20 of 33
- Of value was operational, on the 19 deals whose bridge reconciles
- 63%
On the 19 deals whose bridge closes, operational legs average 63% of the gain and financial legs 37%. Across all 33, the split is 73% operational and 27% financial.
The exit multiple was no safe bet. Only 16 of the 33 deals exited at a higher EBITDA multiple than they entered at. Three of 32 returned less than the equity put in: Freescale, Generac and HD Supply. TXU, later Energy Future Holdings, was a total loss.
What it means for the work after close
If most of the value in these deals came from what the business did, then most of the return was earned after close: in the growth, the margin and the cash the operating plan delivered. That is the work an integration and a value creation plan do, and it is done by people, function by function.
It is also why the value case should not be put away once the price is agreed. We turn the underwrite into value drivers and synergy lines with owners, lock the baseline at Day 1, track each line by workstream and reconcile every one at year end. The board then sees, at any point, whether the deal is still worth what was paid for it. That is our deal value work.
What it does not show
The study compares the model with the filings. It does not compare outcomes with what the sponsors expected, because underwriting cases are not public. Agreement with the filings is evidence that the model's mechanics are right, not a forecast of any deal.
The bridges are not all complete: on 14 deals the entry side does not yet reconcile, for recorded reasons, and 26 of the 33 have at least one figure that no filing discloses. And these are large public buyouts. Most of our work is on deals of up to $1.5bn, where the filings are thinner but, in my experience, the lesson holds: the operating plan carries the return.
Where it started
The method began with my London Business School dissertation, which took the Debenhams buyout apart to see where its return came from. Debenhams is still in the corpus, and still one of the two deals outside the 1% standard, published with its reasons.
The full study lists every deal, every figure and every declared difference.