VC Portfolio Construction: Power-Law Math, Ownership Targets, Reserves + TVPI/DPI Intuition
If you want to sound like a real VC in interviews, this is the topic that separates 'I watched a few podcasts' from 'I can think in fund economics.' VC is not 'pick 30 startups and diversify.' It's power-law math + ownership + follow-on reserves + patience.
Note
Module Reading: This article accompanies the Portfolio Construction & VC Returns module in our Venture Capital interview prep track.
One or two outcomes can drive most (or all) of a fund's returns. That's the defining feature of venture capital, and it shapes everything about how funds are constructed and measured.
This article gives you the fund-level mental model: how VCs think about portfolio construction (shots, ownership, reserves), and how performance is measured (TVPI, DPI, RVPI). Get this right, and you'll outperform most candidates.
1. The VC Power Law (And Why It Changes Everything)
In VC, returns aren't normally distributed. They're extremely skewed: a few deals generate most of the fund's value.
Interview Translation
You're not trying to be "right" 70% of the time. You're trying to (a) access outliers and (b) own enough of them when they happen.
Implications for Portfolio Construction
- Hit rate matters less than hit size. A fund can have many losses and still win if it owns one monster outcome.
- You need a portfolio built to allow outliers. That means enough shots (portfolio size), the right stages, and follow-on capacity.
- Your biggest mistake is "diluting your winners." In power-law land, losing ownership in the one breakout can kill fund-level returns.
Power Law vs Normal Distribution
| Term | Definition | Note |
|---|---|---|
| Normal Distribution | Most outcomes cluster around the mean | Think: hedge fund returns, PE multiples |
| Power Law | A few extreme outliers dominate all others | 1-2 deals can return the entire fund |
| Implication | Portfolio construction is about accessing and owning outliers | Not about diversifying away risk |
2. The "Fund Math" You Need in Interviews
Committed Capital ≠ Investable Capital
Funds have fees and expenses. The standard model is still commonly described as "2 and 20" (management fee + carried interest). Before you can deploy capital, you need to account for the management fee haircut.
Simple Interview Line
"I mentally haircut committed capital for fees/expenses, then split investable capital into initial checks vs reserves."
A Clean Way to Model a Fund in Your Head
Average Initial Check = (Fund Size × (1 - Fee Haircut) × (1 - Reserve Ratio)) / NThis ties portfolio size, check size, and ownership together in one logical chain.
Fund Size=Total committed capitalFee Haircut=~15-20% over fund life for fees/expensesReserve Ratio=% of investable capital saved for follow-ons (often 40-60%)N=Number of initial investmentsWhy this matters in interviews: It shows you understand how check size, portfolio size, and ownership connect. You're not just naming terms—you're thinking like a fund manager.
3. Ownership Targets: The Only Way the Power Law Pays You
Power law only helps if you own enough of the outliers.
Why Pro-Rata Rights Matter
Pro-rata rights let an investor maintain ownership in later rounds by investing proportionally. Without them, your stake shrinks dramatically:
The Dilution Reality
If you start with 15% ownership at seed and don't follow on, you might end up with 3-5% by Series C. That's the difference between a fund-returning investment and a modest win.
A Practical Ownership Framework (Interview-Friendly)
Think in three layers:
Ownership Framework
| Term | Definition | Note |
|---|---|---|
| Entry Ownership | What you want right after your round | Seed: 10-20%, Series A: 15-25% |
| Supported Ownership | What you can keep if you follow pro-rata | Requires reserves and willingness to deploy |
| Exit Ownership | Minimum meaningful stake at exit | What makes the win actually move the fund |
You don't need perfect numbers in interviews. You need the logic: if you expect multiple rounds and option pool refreshes, ownership will dilute unless you reserve capital to follow on.
Common Candidate Mistake
Talking about "finding unicorns" without explaining how the fund keeps ownership as the company compounds.
4. Reserves & Follow-Ons: Defending Winners, Killing Losers
Most VC funds set aside a large share of capital for follow-ons. You'll often hear ranges like 40-60% in practice discussions. Some operators describe reserves as multiples of the initial check (e.g., planning to keep 2-3x the initial to support a winner across rounds).
Three Follow-On Strategies (MECE)
Follow-On Strategy Framework
| Term | Definition | Note |
|---|---|---|
| 1. Pro-Rata Only | Follow mechanically to maintain ownership | Simple, disciplined, doesn't require picking winners again |
| 2. Double-Down | Concentrate more into breakout performers | Embrace power law—winners deserve more capital |
| 3. Triage | Stop funding the 'walking dead' | Recycle attention and reserves to high-signal companies |
Interview One-Liner
"Reserves are how the fund turns early access into meaningful ownership in outliers—either by pro-rata defense or concentrated doubling down."
5. Pacing, Vintage Risk, and Stage Concentration
Even great investors can look dumb if they deploy in the wrong market window.
What Interviewers Want to Hear
- Funds invest over a multi-year investment period, not all at once.
- You want to manage vintage risk (don't put the whole fund into one pricing regime).
- Stage focus matters: seed vs Series A vs growth changes portfolio math, loss rate, check size, and reserve needs.
Stage Strategy Trade-offs
| Aspect | Earlier Stage (Seed) | Later Stage (Growth) |
|---|---|---|
| Portfolio Size | More investments (20-40+) | Fewer investments (10-20) |
| Loss Rate | Higher (many zeros) | Lower (fewer failures) |
| Outlier Potential | Bigger potential multiples | More moderate multiples |
| Ownership | Easier to get large stakes | Harder to 'own big' without huge checks |
| Reserve Needs | More rounds to follow | Fewer rounds to exit |
6. TVPI vs DPI vs RVPI: What Each Really Tells You
These are the most common fund performance multiples you'll be asked about.
Definitions (The Interview-Safe Version)
TVPI = DPI + RVPITotal Value to Paid-In equals Distributions plus Residual Value, all divided by Paid-In Capital.
DPI=Distributions to Paid-In (realized cash returned / capital paid in)RVPI=Residual Value to Paid-In (remaining unrealized value / capital paid in)TVPI=Total Value to Paid-In (total value / capital paid in)Memorable Line
"DPI is reality, RVPI is marks, TVPI is the combined picture."
How to Talk About Them Without Sounding Naive
- Early in a fund's life, DPI can be low (few exits), while RVPI carries the story.
- Later in a fund's life, strong funds convert RVPI → DPI as exits happen.
- A high TVPI with weak DPI can be fine early, but it can also mean you're looking at paper gains that may compress.
Fund Performance Metrics Cheat Sheet
| Term | Definition | Note |
|---|---|---|
| DPI | Cash returned ÷ Capital paid in | The 'truth test'—actual money back to LPs |
| RVPI | Remaining value ÷ Capital paid in | Based on marks—may or may not materialize |
| TVPI | DPI + RVPI | Total picture, but RVPI portion is uncertain |
| IRR | Time-weighted annualized return | Penalizes slow exits, rewards quick distributions |
7. Interview Answers That Score Points
The 30-Second Answer: "How Do VCs Construct Portfolios?"
Model Answer
"VC returns follow a power law, so portfolio construction is about getting enough shots to find outliers, then owning them through reserves and follow-ons. Practically: decide fund size and fee haircut, split capital into initial checks vs reserves, target ownership at entry, protect it via pro-rata, and pace deployment across vintages and stages. Finally, evaluate outcomes using TVPI/DPI/RVPI depending on fund age."
8. Quick Practice Mini-Case (With Solution)
Mini-Case: Fund Construction
A seed fund raises €60M. Assume:
- 50% reserves for follow-ons
- €30M for initial checks
- Target 30 initial investments
Questions:
- What's the average initial check?
- If the fund tries to keep 2x initial available as reserves for winners, how many companies can be "seriously supported"?
- If a winner raises multiple rounds, what tool helps the fund maintain ownership?
Solution
- €30M ÷ 30 = €1.0M average initial check
- If "serious support" ≈ 2x initial, that's ~€2M follow-on capacity per supported winner. With €30M reserves, you can seriously support ~15 companies in theory—but in reality you won't deploy evenly; you'll concentrate into a smaller set of top performers.
- Pro-rata rights (and the decision to exercise them) are the standard mechanism.
Why This Case Works in Interviews
It tests whether you can connect fund size → reserves → check size → portfolio size → ownership in one logical flow. The "2x support" question tests whether you understand that reserves are concentrated into winners, not spread evenly.
9. Frequently Asked Questions
Is TVPI the same as MOIC?
Often they're used similarly as "multiple" measures. TVPI is explicitly total value relative to paid-in capital, and is commonly discussed alongside DPI/RVPI at the fund level. MOIC typically refers to deal-level multiples.
Why can high RVPI be risky?
It relies on marks (unrealized valuation). If pricing compresses or exits disappoint, RVPI can fall before it ever becomes DPI. A fund showing 0.3x DPI and 1.7x RVPI has only returned 30% of capital as cash—the rest is paper.
What's the cleanest way to explain reserves?
"Reserves are capital held back to follow on—either to maintain ownership (pro-rata) or to double down on the few potential outliers."
Where do reserve percentages like 40-60% come from?
They show up frequently in practical fund modeling and follow-on strategy discussions, especially for early-stage funds that expect multiple rounds per winner. The exact ratio depends on stage focus, expected dilution, and fund strategy.
What's the founder-side view of follow-ons?
Founders often hear that investors may reserve additional capital for future rounds and that follow-on participation is tied to strategy, signaling, and ownership defense. When an existing investor doesn't follow on, it can be a negative signal.
Key Takeaways
What to Remember
- Power law drives everything: A few outliers create most fund value—portfolio construction is about accessing and owning them.
- Fund math is one logical chain: Committed capital → fee haircut → investable capital → initial checks + reserves → portfolio size → ownership targets.
- Ownership must be defended: Entry ownership means nothing if you dilute to insignificance—pro-rata rights and reserves are the tools.
- Reserves fund your convictions: Pro-rata to maintain, double-down to concentrate, triage to reallocate from losers.
- TVPI = DPI + RVPI: DPI is realized cash (reality), RVPI is marks (paper), TVPI is the combined picture. Strong funds convert RVPI → DPI.
- Stage changes the math: Earlier = more shots, higher variance, more reserves needed. Later = fewer deals, lower loss rates, harder to own big.
If you can explain power law → ownership → reserves → TVPI/DPI cleanly, you'll outperform most candidates. This is fund-level thinking—the mindset VCs are actually hiring for.