For a decade, private equity returns leaned on borrowed money and rising prices. Neither is available now. What remains is the operating plan, and the operating partner who must deliver it.
With leverage and rising multiples no longer carrying returns, the operating plan has to. The operating partners who succeed will be the ones who can show, using each company’s own data, what the work produced.
For a decade, private equity returns leaned on borrowed money and rising prices. Neither is available now. Entry multiples sit near records, lenders fund about one-third of the purchase price where they once funded half, and the Federal Reserve, after raising rates in September 2026, looks set to raise them further. What remains is the operating plan, and the operating partner who must deliver it.
That leaves three questions about every portfolio company’s growth plan:
Bain’s 2026 private equity report illustrates the shift. A typical buyout in 2015 needed about 5 percent annual EBITDA growth to return 2.5 times its equity in five years. With today’s prices, leverage, and borrowing costs, a comparable deal needs 10 to 12 percent.
Most companies are not growing that fast.
The Golub Capital Altman Index, which tracks private equity-backed middle-market borrowers, shows median earnings growth averaging about 3.5 percent over the past four quarters, from flat in early 2026 to 6.9 percent in the third quarter.
The middle market is not exempt. PitchBook and StepStone data show US buyouts below $1 billion entering at a median 10.8 times EBITDA with 4 times debt in 2025, against 16 times for larger deals. Prices are lower, but so is leverage, and the growth requirement stays high. Sponsors have responded by buying smaller: add-ons made up roughly three-quarters of buyouts in the second quarter of 2026.
Inflation has proven stubborn. In October 2026, core consumer prices rose at an annualized pace near 3.7 percent. The Fed lifted its target range to 3.75 to 4 percent in September, and markets now treat another increase in December as close to certain. The bond market has moved accordingly: after the October inflation report, the two-year Treasury yield jumped to 4.73 percent, and the 10-year sits near 5 percent, within reach of the 24-year high it set in October. We expect the 10-year to hold between 4.75 and 5.25 percent over the next year, so the squeeze falls hardest on floating-rate debt, which is how most buyouts are financed. We see a relatively high likelihood that the Fed raises its target by at least a further half point over the coming three to six months, and we expect the range to peak between 4.50 and 5 percent, 75 to 125 basis points above today’s level. We expect inflation to ease only toward 3 percent by late 2027, which argues against planning on rate relief during the next year.
For portfolio companies, four consequences follow.
At four times debt to EBITDA, each half-point rise in rates costs about 2 percent of EBITDA in added interest, before any hedges. The full move we expect would cost 3 to 5 percent.
Higher rates raise the return buyers require, so a plan that assumes today’s multiple at exit is optimistic. The 10-year Treasury yields about 2.8 percent after inflation, the kind of real return that long-lived assets must now compete with. Every turn of multiple lost has to be replaced with earnings.
With headline inflation running around 4 percent by private estimates, earnings growth of 3.5 percent is roughly flat in real terms. Buyers will ask how much of a company’s growth came from volume and how much from price, and whether price kept pace with wages and input costs.
Our base case is an economy that keeps muddling through: no recession, but no broad acceleration either. Real wages are falling, and we do not expect housing to recover until 2028 or later. Operating plans that count on a rising market deserve a second look.
The tail to watch is stagflation. The two risks we rank highest are inflation that proves stickier, pushing the Fed above 5 percent, and a consumer pullback as real wages fall. Together they would bring higher interest, softer demand, and less pricing power at once. A pullback in AI investment, while less likely, is also a risk worth watching due to AI’s outsized current impulse to the broader US economy.
McKinsey’s 2026 private markets report, using StepStone deal data, finds that for deals done between 2010 and 2022, leverage and multiple expansion supplied 59 percent of returns. Revenue growth and margin improvement supplied 41 percent, and that share did not rise over the period.
Timing matters too. For deals exited since 2019, about 6 percent of ending EBITDA margin arrived in the final year of ownership, against roughly 1 percent in each early year. MSCI finds a similar pattern in its own data: median margins sit below entry levels for much of the hold, then surge in the final six quarters before a sale. It advises acquirers to examine those late gains closely.
The one academic study built on actual value creation plans, by Biesinger, Bircan, and Ljungqvist, covers 1,580 deals. It finds that carrying out the plan raises investor returns beyond what deal selection explains, that operating capacity is scarce, and that the improvements outlast the sponsor’s exit.
McKinsey reports that specialist funds drew 43 percent of their equity value from margin expansion, against 10 percent for generalists.
In FTI Consulting’s 2026 survey of 555 private equity leaders, firms that beat their business case used pricing, sales effectiveness, and customer retention programs nearly twice as often as peers. The results are self-reported, but the pattern is consistent.
A study of 240 buyouts of private US firms by Cohn, Hotchkiss, and Towery, built on tax returns, finds growth through investment and add-on acquisitions, profitability gains concentrated in the weakest companies, and little role for financial engineering. The middle market is where operating work has its clearest record.
McKinsey finds operating groups have more than doubled in size since 2021. No large study links any operating model to returns, and one survey of 267 European buyouts found that operational involvement, unlike strategic and financial involvement, showed no relationship to performance.
OneGuide’s State of Portfolio Ops Report 2026 finds that 51 percent of firms see efficiency gains from AI in most of their portfolio companies, yet fewer than 6 percent see a 10 percent lift in EBITDA or revenue. In FTI’s survey, 19 percent of high performers beat their AI business case, against 5 percent of others. PitchBook’s 2026 survey names data readiness as the top barrier and unclear return as the second. In L.E.K.’s 2026 survey of US buyout professionals, add-on acquisitions led the levers that delivered EBITDA, while AI deployment ranked near the bottom.
Acquisitions rose from the last priority to the first in FTI’s survey, yet only 35 percent of respondents call integration efficient, and health care scores lowest.
Return breakdowns rarely separate organic growth from acquired revenue, or the plan’s effect from a favorable market.
On Engineering Alpha in Private Equity, the podcast Horizon co-hosts, operating leaders describe how they are closing these gaps.
Cory Eaves, Partner and Head of Portfolio Operations at BayPine, describes an “analog to digital” strategy aimed at the roughly 85 percent of the economy outside the technology sector. His firm looks for a “Goldilocks zone”: companies that are neither digital leaders nor digital deserts, but have sound core systems and organized data ready to be put to work. BayPine gave thousands of AI seats to frontline workers, and at one manufacturer, plant staff used them to diagnose machine errors on the spot, cutting downtime from four days to minutes. Eaves frames the gain as redeployment: automate administrative work so people can spend their time on customers.
Jarrad Berman of TZP Group applies a simple rule: every AI initiative must map to revenue or to margin. And a caution from a recent episode applies to buyers and sellers alike: a one-time cost cut labeled as AI is not a durable stream of earnings, and it does not deserve a higher multiple.
The common thread is discipline. Start where the data is ready, and tie each initiative to a line of the income statement.
Our own research on AI adoption points in a similar direction. AI makes other constraints bind harder: management, process, and data. And the companies adopting it fastest were already the best run, which makes their results a poor guide to what everyone else should expect. Our base case for the next 18 months is that AI investment continues while payoffs stay uneven, and that its effect on jobs shows up as a shift in the mix of work, not in headline employment. The main risk to that case is financial: much of the buildout is now funded with debt, and higher rates make that funding harder to sustain.
With rates rising and multiples under pressure, earnings growth is the one dependable source of return, and it has to be demonstrable. The sponsors that can show which initiative moved which number will hold the advantage, and not only with buyers. Lenders are restoring covenants and want reliable reporting. Continuation vehicles, now about 14 percent of sponsor exits by Jefferies’ count, put each company in front of a new and skeptical investor even when the sponsor keeps it. Margin gains that appear late, without evidence of their cause, will be discounted.
EY’s 2026 exit readiness study, a survey of private equity executives and leaders of recently sold companies, makes the point directly. Showing that value creation initiatives produced the EBITDA a buyer is paying for is the hardest part of preparing a sale, and the part that most affects the result. Sixty percent of sponsors call a robust set of data and KPIs their toughest finance issue, and preparation that starts 12 to 24 months ahead pays best.
That evidence has to be built from the first 100 days, not assembled in the last twelve months.
Measurement is only the start. The analytics that pay off are the ones tied to actions an operating partner can take.
Set a baseline at close, then account for each initiative’s effect from market, mix, and acquired revenue. Shift scarce operating attention toward the initiatives that pay or that the broader plan depends on.
Flag the customers most likely to leave, know what to do, and act before renewal; find where price realization lags and correct it; point sales effort at the accounts and markets with the most headroom.
Bring each acquired business onto common definitions within months, so organic growth and cross-sell can be measured and maximized.
Match the right kind of AI to its context. (See Horizon’s perspective on this topic.) Find vendors with proven records not just in AI technical acumen but in process discovery, systems redesign, and client management. There is no magic button to push — the kind of systems redesign required to achieve large absolute returns with AI requires person-to-person engagement, hand-to-hand combat, and a clear vision of what’s at stake.
Keep a running record of what a buyer could verify today, and close the gaps while there is still time to act.
Test each company’s cash flow and covenant headroom against rates 75 to 125 basis points higher, and against a harsher case of rates above 5 percent with flat volume. Track price realization against wage and input costs by product and customer, and favor initiatives that pay back within the year.
The first step is a governed data foundation: clear ownership and stewardship of each data domain, a semantic layer, and shared definitions of customer, product, site, and revenue across every acquired business.
The second step is economics. Attributing a result to an initiative means comparing the customers, sites, or teams that received it with similar ones that did not, and accounting for changes in volume, mix, and price along the way. Those methods turn a claim into causal evidence that a buyer will accept.
With governance in place, a well-built semantic layer, AI tools that work against it, and the right training, finance and operations leaders can answer their own questions, while IT concentrates on keeping systems reliable.
We have done this kind of work before, including analytics and hands-on work with operators that lifted labor productivity 15 percent across about 100 facilities.
The operating partners who succeed in this market will be the ones who can show, using each company’s own data, what the work produced.
A short conversation is enough to see which of these measures your portfolio data can support today, and where the foundation needs work.
An independent read on whether a portfolio company’s value creation story will hold up with buyers, lenders, and continuation-vehicle investors.
Compare NotesA baseline at close, attribution for each initiative, and the governed data that makes the exit record credible.
Build the Record