There are 32,000 companies sitting inside private equity portfolios right now. They are worth a combined $3.8 trillion. And a significant portion of them have no clear path to exit.
That number is the most important number in private equity and the least discussed in board meetings.
The backlog by the numbers
Global buyout exit value reached $717 billion in 2025, a 47% increase year over year. That sounds healthy until you look at volume. Exit count dropped 2% to 1,570. Fewer exits at higher values means the market is clearing the best assets and leaving the rest.
The average holding period at exit has stretched to seven years, up from five to six during the 2010-2021 era. 40% of portfolio companies have been held for five years or longer, up from 29% in 2019.
Distributions to LPs sit at 14% of NAV. That is the lowest since 2008-09. It has been below 15% for four consecutive years, an industry record that nobody wanted to set.
The math on why this matters is straightforward. IRR declines after year seven. Every additional year of holding erodes the return. A company that would have delivered a 2.5x MOIC on a five-year hold delivers something closer to 2.0x at year seven and worse at year eight.
The backlog is not just an inventory problem. It is a return problem.
Why the exits are not happening
Three forces are keeping companies inside portfolios longer than planned.
Seller expectations have not adjusted. The Bain/StepStone GP survey names inflated seller expectations as the number one deal obstacle. Management teams and sponsors anchored to 2021-2022 valuations are discovering that the market has moved. But acknowledging a lower valuation means acknowledging that the value creation thesis did not fully deliver. That is a difficult conversation for everyone involved.
The quality bifurcation is real. “A” assets trade fast at premium valuations. Everything else sits. GF Data’s analysis shows the quality premium, the gap between above-average and average deals, has compressed to roughly 3%. Fewer companies qualify as above average, defined as 10% or more in trailing twelve month revenue growth and 10% or more in trailing twelve month EBITDA margins. The bar has risen while performance has not kept pace.
Buyers are more rigorous. 80% of GPs expect multiples to remain flat in 2026. When buyers cannot count on multiple expansion to generate returns, they become far more selective about what they buy. Due diligence is slower, deeper, and more data-driven. The screening process that used to take two weeks now takes two days with AI tools, but the diligence itself is more thorough than ever. Inconsistencies that would have been negotiated away in 2021 now kill deals.
The holding period trap
The typical value creation plan is built for a five-year hold. The major operational initiatives are loaded into years one through three. Year four is optimization and harvest. Year five is exit preparation.
When the hold stretches to seven years, the plan breaks in specific ways.
The management team that was hired or incentivized for the original timeline starts to disengage. The operating partner cycles their attention to newer investments. The value creation plan stops being updated because the original milestones have either been hit, missed, or abandoned.
Years six and seven become maintenance years. The company is running, but it is not improving. The EBITDA might be growing at 3-4% instead of the 10-12% the original thesis required. The equity story for exit has not changed in two years because nobody has invested in changing it.
And then the exit process starts. The management team dusts off the board deck. The operating partner reviews the value creation plan for the first time in a year. And the data room needs to be assembled from systems that have not been standardized, metrics that have not been reconciled, and narratives that do not match the underlying numbers.
This is when the cost of the backlog becomes concrete. Not in abstract IRR erosion. In the actual work of trying to sell a company whose data cannot support the equity story.
What the best assets have in common
The companies that are exiting in this environment share characteristics that have nothing to do with sector or size.
They can answer buyer questions quickly. When the diligence team asks for MRR by customer segment reconciled to the general ledger for the last 36 months, the answer comes in 48 hours. Not three weeks of manual reconciliation.
Their numbers agree with each other. The CRM revenue matches finance. The customer count in the board deck matches the billing system. The EBITDA bridge has supporting documentation at every step.
Their value creation story is provable. When the equity story says “we improved pricing across the portfolio and gained 200 basis points of margin,” the data shows exactly which customers, which products, and which periods. The buyer can verify it independently.
This is not about having the best BI tool or the most sophisticated analytics platform. It is about having clean, governed, defensible data that lets the buyer build confidence quickly.
The companies without this sit in the backlog. The companies with it trade.
The compounding cost of waiting
Here is where the timing matters most.
If you invest in data infrastructure in year one of the hold and it enables a 2% margin improvement by year two, that improvement compounds through every subsequent year. On a five-year hold, you get four years of compounded benefit. The data work pays for itself and then some.
If you wait until year three, you get two years of compounded benefit on a five-year hold. Still valuable, but half the return.
If you wait until year five, which is what most companies do, you are running data remediation while simultaneously preparing for exit. The data work is now defensive, not offensive. You are not building equity value. You are trying to prevent it from being discounted.
And if the hold stretches to seven years because the exit market is tight? The companies that started data work in year one have six years of compounded improvement. The companies that started in year five have two years of rushed cleanup and a data room that still shows the gaps.
One mid-market platform company failed to establish data integration standards across its add-on acquisitions. The result was a six-month delay in value creation initiatives and $8 million in forgone EBITDA. That delay did not just cost $8 million. It compressed the remaining window for improvement, weakened the exit narrative, and contributed to a lower final multiple.
What this means for portfolios right now
If you are an operating partner with companies in years four through seven of a hold, the instinct is to focus on the exit process. Get the banker. Build the CIM. Start the data room.
That instinct is wrong if the underlying data is not ready.
Starting the exit process with unreliable data does not speed up the timeline. It extends it. Buyers find the inconsistencies. Diligence stalls. Questions generate more questions. The bid adjusts downward or the deal structure shifts from a clean close to an earnout.
The right sequence is different. Fix the data first. Get the metrics reconciled. Establish canonical sources of truth. Make sure the numbers in the board deck match the numbers in the systems. Then start the exit process.
This takes a quarter, not a year. It is not a data transformation. It is targeted work on the specific metrics that buyers will scrutinize.
GF Data’s analysis of 360 mid-market transactions since Q3 2024 shows the payoff directly. Sellers who paired a quality-of-earnings analysis with a data quality assessment achieved 7.4x EBITDA multiples. Sellers without the data quality component achieved 7.0x. That is $2 million on a $50 million business.
The backlog will clear. The question is how.
The $3.8 trillion in unrealized value will not sit forever. LPs are applying pressure. GP fundraising is declining. Management fees are compressing. The system is forcing exits.
The question is whether your portfolio companies exit on the right side of the bifurcation or the wrong one.
The companies that trade at premium multiples and close cleanly will be the ones whose data supports the equity story. The companies that sit, discount, or get restructured into continuation vehicles will be the ones whose data cannot survive scrutiny.
32,000 companies are waiting to exit. The ones that invested in data readiness early will exit first and exit better. The ones that deferred it are discovering that the backlog is not just a market problem. It is a preparation problem.
And in a market where 12 is the new 5, preparation is the entire game.