Serial Acquisition Strategy: How to Price the Growth You Buy

Why a growth rate is a purchase order

A plan says profit will grow at some rate. The businesses you already own will produce part of that. The difference has to be bought, with money, from owners who do not have to sell, and the size of that bill is fixed the day the rate is announced, before anyone has looked at a single target.

Almost no acquirer states that bill, because a growth rate does not look like a purchase order. In the worked case used in the Acquisition Machine Strategy Pack, a plan of nine percent growth against three percent organic commits the company to buying $74.4 million of profit a year at a cost of $818.4 million.

The four numbers a serial acquirer should be able to state

A serial acquisition strategy reduces to four figures the company already has or can produce in a week. When any one of them is missing, the machine is running on a sentence.

The spread is settled entirely by whether the multiple you pay sits below the one you trade at. Where your currency is expensive and you buy cheaply, the spread is real money. The two questions it never asks are how long the balance sheet funds it and what happens when your own multiple falls.

What you pay with: shares, cash or debt

Cisco once paid in its own stock. In September 1993 it handed over two million shares, worth about $97 million at that day's close, for Crescendo Communications, because customers wanted switching gear Cisco did not make. Thirty-one years later it closed its largest deal ever in cash: $157 a share for Splunk, about $28 billion, at a 31% premium to Splunk's unaffected price, more than four times the size of Cerent or Scientific Atlanta. The same acquirer moved from paying with a rich currency to paying with the balance sheet.

Exxon paid for Pioneer entirely in its own shares. The agreement signed on October 10, 2023 gave 2.3234 Exxon shares for each Pioneer share, $59.5 billion in equity and about $64.5 billion with debt, at an 18% premium to Pioneer's undisturbed close and only a 9% premium to its prior 30-day average price. Paying in stock shared the commodity risk with Pioneer's holders. Exxon has since raised its annual synergy target from about $2 billion to $4 billion.

Salesforce built its platform the same way, check by check: about $2.5 billion for ExactTarget in 2013, $2.8 billion for Demandware, $6.5 billion for MuleSoft, $15.7 billion for Tableau and $27.7 billion for Slack. The Salesforce case reads the currency plainly: CRM gave it "a public valuation rich enough to make $15 billion and $27 billion deals payable in stock." A machine funded in shares depends on the one input no acquirer controls, its own multiple.

What you pay for the earnings

A price without a denominator cannot be compared with another deal, with a buyback or with your own record. Broadcom stated its denominator. When it agreed on May 26, 2022 to buy VMware for about $61 billion in cash and stock, plus about $8 billion of assumed net debt, it targeted adding about $8.5 billion of pro forma EBITDA within three years of closing. Broadcom's adjusted EBITDA reached a record $31.9 billion in fiscal 2024, up 37%, and it sold VMware's End-User Computing division to KKR for about $4 billion.

Kraft Heinz shows what happens when the earnings are not what the price assumed. In 2015 Berkshire Hathaway and 3G put in an additional $10 billion to fund the special dividend to Kraft shareholders. In the fourth quarter of 2018 Kraft Heinz reported a $15.4 billion goodwill impairment, and Berkshire took a $3 billion write-down on its investment. Warren Buffett said in February 2019, "We overpaid for Kraft," noting the business used about $7 billion in tangible assets while Berkshire paid roughly $100 billion in total asset value. By September 2025 Berkshire's 27.5% stake, which cost $9.8 billion, was worth about $8.9 billion.

Salesforce paid $27.7 billion for Slack in stock and cash, a 54% premium over Slack's roughly $18 billion market value, on the thesis that Slack would become the new interface for Salesforce. Its case finds the thesis unproven. The two earlier platform deals are easier to read: in fiscal 2024 the segment holding MuleSoft and Tableau produced $5,189 million of subscription and support revenue, about 16% of total subscription and support revenue.

Where the debt stops

A machine funded with debt has a stop date, and it is set by the balance sheet rather than by the pipeline. AB InBev arranged a record $75 billion bridge in November 2015 to buy SABMiller, a deal its own filing valued at roughly £79 billion. By January 27, 2016, after roughly $47 billion of bond issuances, it had canceled $42.5 billion of the facility. Net debt was still $67.6 billion at the end of 2023, down from $69.7 billion, at 3.38x normalized EBITDA against a stated optimal target of about 2x. Moody's and S&P upgraded AB InBev in early 2024, citing the deleveraging.

Dell agreed on October 12, 2015 to buy EMC in a transaction valued at about $67 billion, with banks committing up to $49.5 billion of debt financing. The years after were spent paying for it: Dell sold Dell Services, the Dell Software Group and an enterprise content division for roughly $7 billion aimed at the merger debt, and the 2021 spin-off of VMware came with an $11.5 to $12.0 billion special cash dividend that flowed largely to Dell. A debt-funded deal can still be the right move, but it ends the machine for as long as the debt takes to retire.

What you can read before you sign

Some of the cost of a deal is on the record before it closes. Bayer closed its acquisition of Monsanto on June 7, 2018 at $128 a share, about $63 billion including Monsanto's debt. IARC had classified glyphosate as a probable human carcinogen in March 2015, and Monsanto was already defending over 120 toxic-tort cases. Bayer lost roughly 40% of its market capitalization after closing; 55.5% of shareholders voted against ratifying management's actions at the 2019 annual meeting; and in June 2020 Bayer announced $10.1 billion to $10.9 billion to resolve Roundup litigation.

The opposite case is a price set by someone else. UBS agreed on March 19, 2023 to absorb Credit Suisse in an all-share deal for total consideration of CHF 3 billion, with over CHF 100 billion of liquidity from the Swiss National Bank and a government loss guarantee of up to CHF 9 billion. UBS booked final negative goodwill of USD 27.7 billion. Its chairman's framing was exact: "This acquisition is attractive for UBS shareholders but, let us be clear, as far as Credit Suisse is concerned, this is an emergency rescue." On October 1, 2025 a Swiss court ruled the write-off of Credit Suisse's AT1 bonds unlawful, a cost no spread calculation would have priced.

What these cases show

Each case is drawn from the Stratrix case library and checked against the filings and reporting listed under Sources. The last column says what each one does not establish.

Running the decision

What this method cannot tell you

It prices the machine, not any deal in it. Integration, synergies, the people who leave and what the acquired businesses do afterward are real, often decisive, and outside this arithmetic. No disclosure separates the effect of a multiple from everything else moving in the same period, and none of the cases above is presented as doing so.

It also cannot forecast your own multiple, which is the input the whole engine runs on and the one no one controls. The cases support recognizing a mechanism; they cannot support a success rate for serial acquirers, and this page does not claim one.

Questions