The proof of value creation: showing where the EBITDA growth came from

EBITDA at entry was €4.0 million. Three years later it is €6.0 million. The slide in the fund's annual review shows a fifty percent increase.

The question that follows is not how much EBITDA grew. It is where the growth came from. How much arrived with the company that was acquired in year two. How much is price, and how much of that price only covered higher costs. How much is an item that used to count against EBITDA and now does not.

A fund that cannot answer still has the growth. What it has lost is the explanation, which a buyer's adviser will now write in its place.

Why the question is sharper now

Bain & Company's 2026 Global Private Equity Report gives an illustration. A typical buyout of 2015 needed about 5 percent annual EBITDA growth to return 2.5 times invested capital over five years. A typical buyout today needs something closer to 10 to 12 percent. Entry prices are higher, debt costs more and is used less, and the exit multiple no longer rises by itself. Bain's example is American.

An older European study shows what used to do the work. Achleitner and her co-authors examined 206 realised European buyouts and attributed one third of sponsor returns to leverage, and two thirds to operational and market effects together. With leverage dearer and the market flat, the operating part has to carry more of the return.

So the growth in EBITDA has to be real, and the fund has to be able to show which part of it is.

Three kinds of growth

Bought

An add-on acquisition brings its EBITDA with it. This is real value, and it can be good value: EBITDA bought at five times and later sold at seven is a recognised way to make money. But it was made with capital. It says nothing about whether the original business improved.

Earned

This is what the investment case promised for the business the fund first bought: price, volume, mix and cost. It is the hardest part to measure, because the effects offset each other. A price increase that only passes on higher costs raises revenue and adds nothing to EBITDA.

Redefined

The third kind is a change in what is counted. A cost is treated as non-recurring and added back. Some of these costs really were one-off. Some come back every year under a new description. In neither case did the business change. The first kind, if it can be proved, does say something about what the business earns. The second says nothing.

One company, used throughout this series

The company below is invented. We use the same company in all five articles of this series, seen at one date, 31 March 2026. Its numbers are chosen to add up, and they describe no client. The fund bought it three years earlier, at seven times its EBITDA of €4.0 million, with €12.0 million of debt and €16.0 million of equity.

The bridge is drawn on a fixed perimeter. The add-on is shown as one line, at the EBITDA it contributes today, including whatever it has gained or lost since it was bought. Every other line belongs to the original business.

Of the €2.0 million increase, €0.7 million was bought and €0.6 million is adjustments that did not exist at entry. The earned growth is the remaining €0.7 million.

Three readings follow.

Price added €1.2 million. Inflation in inputs, wages and overheads took €1.3 million. The company raised prices for three years and did not quite keep pace with its costs. Its earned growth came from volume and from savings.

The headline growth is about 14 percent a year. The earned growth, €0.7 million on €4.0 million over three years, is just over 5 percent a year. Both figures are true. The second is what the fund can prove about the business it bought.

It may be too low. If the €0.4 million of one-off costs is what the company says it is, the business earned €1.1 million, about 8 percent a year. Without the documents nobody can show that, and a buyer will not assume it.

The add-on was bought for €3.5 million, five times the EBITDA it had then and still has, and was paid for with debt. Net debt was €12.0 million at entry and is €14.0 million today. So since entry, trading has repaid €1.5 million.

That last figure deserves a second look. Over three years the company reported about €14 million of EBITDA. Interest took roughly €2.2 million and tax €2.3 million. Capital spending took €5.0 million. Growth in working capital took €3.0 million, twice what reached the lenders, and that is the subject of the next article in this series. What remained for the lenders was €1.5 million. A bridge of EBITDA says nothing about any of this. A fund that reads only the bridge will believe the company is producing cash at a rate it is not.

None of this makes the investment a bad one. It changes what the fund can truthfully say about it.

What makes the bridge hard to build in Greece

The arithmetic is simple. The records are the difficulty.

The Greek Accounting Standards, Law 4308/2014, set a chart of accounts that records expenses by nature. Employee benefits are in account 60, sundry operating expenses in 64, depreciation in 66, before anything is transferred to cost centres or products. The income statement may then be presented by function or by nature. Either way, the split between cost of sales and overheads is an allocation made after the ledger. If the allocation keys change between years, gross margin moves when the business has not.

Price and volume cannot be separated in a ledger of values. Quantities are held in the invoicing or warehouse system, at invoice-line level. If product codes were reorganised since entry, three years of quantities no longer line up with three years of values.

EBITDA is not a line in the statutory formats. Each company defines its own, and the definition is seldom written down. Adjustments are agreed at a board meeting and applied in the pack. Three years later nobody can list them with their amounts.

A fund that wants to follow the bridge monthly meets one more obstacle. Private-sector employees in Greece receive a Christmas bonus of one monthly salary, an Easter bonus of half, and a leave allowance of up to half. Where these are expensed when paid, some months carry extra payroll and the rest carry none. The annual figure is right, but single months cannot be compared.

A layout: the adjustments register

One table, kept from the day of the investment, settles most arguments about add-backs. This is a layout. The row shown is invented.

Each row needs its document. If the document is missing, the row is only a claim. The last column does different work. It catches the cost that is called one-off and returns the next year. No document rescues that one, and the register's job is to show it to the fund before a buyer points it out.

Four questions for a portfolio company

1. Can the company split EBITDA growth since entry into bought, earned and redefined, and reconcile the total to the audited accounts?
2. For its largest product or customer groups, has price grown faster than cost since entry, shown with quantities and not only values?
3. Is there a list of every EBITDA adjustment since entry, with the amount in each year and the document behind it?
4. What did each add-on cost, how was it paid for, and what has trading done to net debt since?

A first answer to each takes days. It will be approximate, and it will show how far the company is from an exact one. Where quantities and a stable cost mapping do not exist, the exact answer cannot be rebuilt afterwards. It has to be recorded as the months go by.

A company whose finance team already closes each month on fixed definitions and keeps this register can answer all four today. It does not need outside help with this.

Sources

• Bain & Company (2026). Global Private Equity Report 2026, section "12 is the new 5". Illustrative United States example: about 5 percent annual EBITDA growth for a 2.5 times return on a 2015 buyout, against roughly 10 to 12 percent today.
• Achleitner, A.-K., Braun, R., Engel, N., Figge, C. & Tappeiner, F. (2010). Value Creation Drivers in Private Equity Buyouts: Empirical Evidence from Europe. The Journal of Private Equity, 13(2), 17–27. Sample of 206 realised European buyouts; one third of sponsor returns attributed to leverage, two thirds to operational and market effects.
• Law 4308/2014, Greek Accounting Standards. Annex C, chart of accounts, group 6; article 16, income statement formats by function and by nature.
• Joint Ministerial Decision 19040/1981 (Christmas and Easter bonuses); Law 4504/1966, article 3 (leave allowance).