What Loss Ratio Can a Venture Portfolio Sustain and Still Produce a 3x Return?
The NVCA 2026 Yearbook illustrates the outlier-heavy nature of recent venture markets. A small number of very large financings can dominate aggregate activity. Venture fund return models should be honest about dependence on large winners.
NVCA reported 487 mega-deals of $100M+, equal to 3.2% of deal count and 67% of total deal value in 2025.
Capital Loss Matters More Than Company Count
A venture fund can lose more than half of its companies and still produce a strong return. It cannot lose most of its capital and expect the same answer. The useful loss ratio is therefore capital weighted: how much money went into the companies that failed, including follow-ons, not simply how many names went to zero.
In the simplified $100 million model, a 40% capital loss leaves $60 million to create $300 million of value. The surviving capital must return 5.0x. If 60% of capital is lost, the surviving $40 million must return 7.5x. At a 75% loss, the requirement rises to 12.0x. Those are portfolio-level results before any extra gap between gross and net returns.
Why Company Count Can Mislead
Imagine two funds that each lose six of ten companies. Fund A made small first cheques in the six losses and put most of its follow-on capital into the four winners. Fund B kept supporting weak companies and lost much larger amounts in the same six names. Both report a 60% company loss rate, but their return maths are completely different.
This is why an LP should ask for cost, current value, and proceeds by company. The question is not whether the manager was wrong six times. It is how much the mistakes cost, whether the manager stopped adding capital when the facts changed, and whether enough ownership remained in the winners.
Three Ways to Read a Loss Ratio
| Measure | What it counts | What it tells an LP | What it can hide |
|---|---|---|---|
| Company-count loss | Number of companies below cost or at zero | How often the manager's selections failed | Whether the losses were small or heavily funded |
| Capital-weighted loss | Share of invested dollars that produced no return | How much damage the losses did to the fund | Timing and any small recoveries |
| Loss after reserves | Initial and follow-on capital lost in each company | Whether later decisions improved or worsened the result | Opportunity cost of not backing a winner |
To produce three times the original capital, the surviving capital must return three times when there are no losses, five times when forty percent of capital is lost, seven point five times when sixty percent is lost, twelve times when seventy five percent is lost, and fifteen times when eighty percent is lost.
Losses Make the Required Winner Multiple Rise Rapidly
Once most invested capital is lost, the surviving companies need unusually large outcomes to preserve a 3x portfolio result.
View sensitivity data and formula
| Capital loss ratio | Capital remaining | Multiple needed on remaining capital |
|---|---|---|
| 0% | 100% | 3.0x |
| 20% | 80% | 3.75x |
| 40% | 60% | 5.0x |
| 60% | 40% | 7.5x |
| 75% | 25% | 12.0x |
| 80% | 20% | 15.0x |
Reserves Matter More Than the First Cheque
Venture losses often become expensive in later rounds. A small opening cheque may be reasonable when little is known. The harder decision comes after the company misses a plan, needs more time, and offers existing investors the chance to protect their stake. Supporting it can preserve upside; it can also turn a controlled loss into a major one.
The available evidence shows why this matters. Has customer retention improved? Is the next round led by an investor with an independent view? Does the company have enough cash to reach a clear milestone? Is the price fair? A manager should be able to explain why more capital went into a company and what would have caused the fund to stop.
A 3x Target Still Depends on the Winners
Reducing losses does not by itself create a 3x fund. A portfolio of modest 1.5x and 2.0x outcomes may avoid zeros and still fall well short after fees and carry. Venture works when some companies create returns that are large compared with the size of the fund. LPs should test the portfolio both ways. First, remove the largest winner and see what remains. Then cap the top few outcomes at more ordinary exit values. If the model collapses, that does not automatically make it bad, but it shows how much the result depends on rare outcomes and on the fund keeping enough ownership in them.
Questions for the Investment Memo
- Show both company count and invested capital.
- Separate initial and follow-on cost: This reveals whether reserve decisions added value.
- Show partial recoveries: A 0.3x return is not a zero, but it still leaves a large hole.
- Test ownership in the winners: A large company outcome may not translate into a large fund outcome after dilution.
- Use net fund targets: Company-level multiples should be carried through fees, expenses, and carry.
Define Loss Ratio Clearly
The numbers make the effect easier to see. In a 20-company portfolio, 12 complete losses equal a 60% company-count loss ratio. That says nothing about whether the lost companies consumed 30% or 70% of capital.
NVCA reported that 3.2% of 2025 deals represented 67% of total US VC deal value. That kind of skew is why winner magnitude matters more than average outcome.
What the Winners Must Do
A $100 million fund targeting $300 million of net value can lose $40 million of invested capital and still work if the remaining $60 million produces $300 million. That requires a 5.0x net multiple on the surviving capital before allowing for any additional fund-level leakage.
On a $100 million fund targeting $300 million of net value, losing 40%, 60%, or 75% of invested capital leaves $60 million, $40 million, or $25 million of surviving capital. Those surviving positions must produce net multiples of 5.0x, 7.5x, or 12.0x respectively.
Loss Ratio and Winner Requirement
The $300M net target stays fixed, so higher losses force a much larger multiple from the surviving capital.
View scenario data and assumptions
| Capital-loss scenario | Capital lost | Surviving capital | Net value target | Required multiple on surviving capital |
|---|---|---|---|---|
| 40% loss | $40M | $60M | $300M | 5.0x |
| 60% loss | $60M | $40M | $300M | 7.5x |
| 75% loss | $75M | $25M | $300M | 12.0x |
Frequently Asked Questions
Can a fund lose more than half its companies and still be strong?
Yes: Venture returns can be power-law driven. The question is whether the winners can return enough capital after dilution and fund costs.
Should managers avoid losses at all costs?
No: Avoiding all loss can mean avoiding venture risk. The better goal is to size risk appropriately and reserve for the companies with evidence.