Why are these businesses valued on different measures at all?
Because buyers price what they believe is durable. Where earnings are the durable thing, the market pays a multiple of earnings. Where the earnings are small, absent or deliberately suppressed in favour of growth, the market pays a multiple of revenue and takes a view on what the earnings will eventually be.
The comparison most often quoted illustrates the point and misstates it at the same time. As at January 2026, system and application software traded at about 11.41 times enterprise value to revenue while oil and gas production and exploration traded at about 6.21 times enterprise value to EBITDA (Eqvista, published April 2026, citing the January 2026 Damodaran dataset). Those are different denominators. A business earning a 20 percent margin at 11.41 times revenue is being valued at more than 50 times earnings, which is the actual size of the gap.
On a like-for-like transaction basis the spread is real but much smaller. Energy and materials transactions cleared at 7.4 to 8.9 times enterprise value to EBITDA in mid-2026 while information technology transactions cleared at a median of about 12.5 to 12.8 times, against a global median of 10.7 times, the highest since 2021 (IB Interview Questions, June 2026).
So the honest version of the folk claim is that narrative sectors trade at roughly four to five turns of EBITDA above asset-heavy ones, not at the two to three times gap the revenue-multiple comparison appears to show.
| Measure | Narrative sector | Asset-heavy sector |
|---|---|---|
| Enterprise value to revenue, public listed | System and application software at about 11.41 times as at January 2026, with median public software at 3.4 times by March 2026 | Not the measure used, because earnings are the durable quantity |
| Enterprise value to EBITDA, public listed | Aggregate software index at about 26.6 times | Oil and gas production and exploration at about 6.21 times as at January 2026 |
| Enterprise value to EBITDA, transactions | Information technology at a median of about 12.5 to 12.8 times in mid-2026 | Energy and materials at 7.4 to 8.9 times in mid-2026, against a global median of 10.7 times |
What happened to the narrative multiple?
It compressed, and then the measure itself changed. Median public software enterprise value to revenue fell from about 5.1 times at the end of 2025 to 3.4 times by March 2026, while the aggregate software index continued to trade at roughly 26.6 times enterprise value to EBITDA (Aventis Advisors, updated June 2026).
The second half of that is the more consequential development. For the first time since 2015, enterprise value to EBITDA is displacing enterprise value to revenue as the primary valuation measure in software (Aventis Advisors, June 2026). The market has not abandoned the sector. It has changed what it is willing to pay for.
Private marks are following, with the usual lag. Software valuations in private equity portfolios declined about 8 percent in the first quarter of 2026, with the United States at 8.9 percent and Europe at 4.2 percent (Bain and Company, Private Equity Midyear Report 2026, 8 June 2026, on MSCI analysis as at 31 March 2026). A single-digit private adjustment against a far larger public move implies more to come.
For an owner whose business has been valued on a revenue multiple in conversation, that shift is the whole story. The same business, measured on earnings, is a different asset with a different price, and the change happened in the market rather than in the company.
“Owners choose which multiple they are building toward long before they realise they are choosing. Suppress margin to buy growth and you are asking to be valued on revenue by a market that may not be paying revenue multiples when you arrive. That is a bet on the frame, not on the business, and it is worth making deliberately rather than by default.”
Does the financing market agree?
It agrees more emphatically than the equity market does, and it moved first. Software's share of broadly syndicated loan issuance fell to 8.6 percent year to date in 2026 from 17.6 percent in 2025, its lowest since 2013, while healthcare became the largest sector for institutional loan issuance for the first time since 2015 at a record 12.4 percent (PitchBook LCD, as at 30 June 2026; Sikich, 13 July 2026).
Pricing carries the same message. Software credits are quoted at 75 to 100 basis points above the standard matrix by one adviser and at 150 to 300 basis points above comparable non-software credits by another, with interest-only periods cut from three years to two or less and liquidity minimums and cash controls now standard (Houlihan Lokey, Q2 2026; SPP Capital Partners, Market At A Glance, July 2026).
The structural readings behind that are severe. Around 30 percent of outstanding software loans mature by 2028 against 22 percent for the broad leveraged loan market, and software recoveries in private credit are estimated at around one third of par given collateral that is largely intangible (Allianz Research, 20 February 2026, published before the second-quarter 2026 market shock and labelled accordingly). The 2028 software maturity wall of about 40 billion dollars has barely moved through the first half of 2026 (PitchBook LCD, through 30 June 2026).
One practitioner reading is worth quoting for its framing rather than its numbers: the disruption of legacy software business models increasingly looks like a structural credit story rather than a cyclical one (Sikich, 13 July 2026). If that is right, financing availability rather than valuation opinion is the binding constraint for those businesses into 2027.
What should an owner take from this?
That the exit multiple belongs in the plan from the beginning, and that it is a choice with consequences rather than a result. A business run for cash generation is valued on cash generation in every market. A business run for growth at the expense of margin is valued on a frame that can be withdrawn, as it partly was between December 2025 and March 2026.
That does not make the growth path wrong. It makes it a position that has to be exited at some point, and the willingness to sell into enthusiasm rather than through it is what converts a narrative multiple into money. Enthusiasm is not a permanent feature of any sector, and the record shows it lasting years rather than decades.
The counterpart is that asset-heavy and cash-generative businesses are being valued more consistently than their owners typically believe. Global median transaction multiples at 10.7 times, the highest since 2021 (IB Interview Questions, June 2026), are not a hostile environment for a business with real earnings, and the current market rewards exactly the characteristic these businesses have: a return that comes from operations rather than from re-rating.
Underneath the sector question there is a growth question that applies to everyone. Median EBITDA growth across private credit borrowers fell to 24 percent from 27 percent, the largest quarter-on-quarter decline on record, and median revenue growth to 12 percent from 13 percent (KBRA, Q2 2026 Middle Market Compendium, published 28 July 2026, covering the twelve months to 30 June 2026). Growth is getting harder to produce in both frames.
What does not change?
The cash. Whatever measure is fashionable, the buyer is eventually acquiring a stream of money, and every frame is an attempt to guess its size and durability. Revenue multiples are a guess about future earnings. Earnings multiples are a guess about their persistence. Neither is a different asset.
That is why the preparation work is identical across both. Thirty-six months of clean, normalized monthly financial statements, a quality of earnings report commissioned early, and evidence of retention and margin will move price in a software business and in an industrial one, because in both cases they resolve the question the multiple is standing in for (Capstone Partners, Capital Markets Update, 4 June 2026).
It is also why the two-speed market keeps producing the same pattern. High-quality issuers are seeing among the most aggressive pricing and terms in years while marginal credits face rising prices and shrinking leverage (SPP Capital Partners, July 2026), and on the equity side multiples for A-grade targets are very high while lower-grade companies are not attracting bids (ACG and GF Data, Q3 2026 Market Pulse Survey, published 15 July 2026).
The conclusion for an owner is unfashionable and durable. Build the cash, document it, and treat the multiple as the market's opinion of your evidence rather than as a property of your sector.
As of August 2026
Sources: Eqvista, published April 2026, citing the January 2026 Damodaran dataset, for system and application software at about 11.41 times enterprise value to revenue and oil and gas production and exploration at about 6.21 times enterprise value to EBITDA; Aventis Advisors, updated June 2026, for median public software enterprise value to revenue falling from about 5.1 times at the end of 2025 to 3.4 times by March 2026, for the aggregate software index at roughly 26.6 times enterprise value to EBITDA, and for enterprise value to EBITDA displacing enterprise value to revenue as the primary software valuation measure for the first time since 2015; IB Interview Questions, June 2026, for energy and materials transactions at 7.4 to 8.9 times, information technology transactions at a median of about 12.5 to 12.8 times, and a global median transaction multiple of 10.7 times, the highest since 2021; Bain and Company, Private Equity Midyear Report 2026, 8 June 2026, on MSCI analysis as at 31 March 2026, for software valuations in private equity portfolios declining about 8 percent in the first quarter of 2026, with the United States at 8.9 percent and Europe at 4.2 percent; PitchBook LCD, as at 30 June 2026, for software falling to 8.6 percent of broadly syndicated loan issuance year to date from 17.6 percent in 2025 and for the 2028 software maturity wall of about 40 billion dollars barely moving through the first half of 2026; Sikich, 13 July 2026, for healthcare becoming the largest sector for institutional loan issuance at a record 12.4 percent and for the characterisation of software credit as a structural rather than cyclical story; Houlihan Lokey, Q2 2026, and SPP Capital Partners, Market At A Glance, July 2026, for software pricing premiums, shortened interest-only periods and standard liquidity controls, and for the two-speed credit market; Allianz Research, 20 February 2026, for around 30 percent of outstanding software loans maturing by 2028 against 22 percent for the broad leveraged loan market and for software recoveries estimated at around one third of par, labelled as predating the second-quarter 2026 market shock; KBRA, Q2 2026 Middle Market Compendium, published 28 July 2026, covering the twelve months to 30 June 2026, for median EBITDA growth falling to 24 percent from 27 percent and revenue growth to 12 percent from 13 percent; ACG and GF Data, Q3 2026 Market Pulse Survey, published 15 July 2026, for the split between A-grade and lower-grade targets; Capstone Partners, Capital Markets Update, 4 June 2026, for the preparation standard.