TVPI

Imagine you gave someone ¥100 to invest for you.

They've already given you back ¥50. And the investments they're still holding for you are worth ¥80.

TVPI asks: "How much is everything worth compared to what I gave you?"

¥50 (what you got back) + ¥80 (what's still growing) = ¥130 total value.

¥130 ÷ ¥100 = 1.3x TVPI. Your money grew by 30%!

TVPI stands for "Total Value to Paid-In Capital." It's the most common way to measure how well a VC or private equity fund is doing.

Think of it as two parts combined:

• Distributions: Money the fund has already returned to investors (from selling companies, dividends, etc.)
• Remaining Value: What the unsold investments are currently worth

A 2.0x TVPI means LPs doubled their money. Top-tier VC funds aim for 2.5x or higher.

TVPI breaks down into two components that tell different stories:

• DPI (Distributions to Paid-In): The "realized" part — actual cash returned. This is money in the bank.
• RVPI (Residual Value to Paid-In): The "unrealized" part — paper gains from companies not yet sold.

Why it matters: RVPI is based on the GP's valuation of portfolio companies. Until there's an exit, it's an estimate. Fund A's 2.0x is much more "real" than Fund B's.

The J-curve: Young funds typically show low TVPI because they've deployed capital but companies haven't matured. TVPI improves as exits happen — usually starting around years 4-5.

Gross vs. Net TVPI:

• Gross TVPI: Before fees and carry — what the portfolio generated
• Net TVPI: After management fees (typically 2%) and carried interest (typically 20%) — what LPs actually receive

A 3.0x gross might translate to ~2.3x net. Always compare net-to-net.

The valuation problem:

RVPI depends entirely on how GPs mark their portfolios. Private companies don't have market prices, so fair value is estimated. This creates room for "optimism" — especially before fundraising for the next fund.

Common markup approaches:

• Last round valuation (most common)
• Comparable public company multiples
• Discounted cash flow (rare for early-stage)

Industry benchmarks (VC funds, net TVPI):

• Top quartile: >2.5x
• Median: 1.5-2.0x
• Bottom quartile: <1.5x

Vintage year matters: 2010-2014 vintages look spectacular due to massive tech exits. 2021 vintages are underwater. Always compare same-vintage quartiles using Cambridge Associates, Burgiss, or Preqin data.

TVPI is NAV-sensitive and vintage-dependent. Sophisticated LPs treat it as one data point, not gospel.

The denominator games:

• Some GPs exclude recycled capital from "Paid-In" — inflating TVPI
• Management fee offsets can distort the denominator
• Subscription lines delay capital calls, affecting both timing and reported multiples

GP-led secondaries and continuation funds:

When a GP rolls a company into a new vehicle at a GP-set valuation, it crystallizes an "exit" for TVPI purposes. But was that price market-tested? Some LPs discount these "exits" when evaluating true DPI.

NAV lending distortions:

Funds increasingly borrow against portfolio NAV to fund distributions. This boosts DPI (cash out) without actual exits, making TVPI look more "realized." But leverage adds risk — if marks drop, the fund may face margin calls.

Cross-vintage and cross-strategy comparison:

Comparing TVPI across vintages requires adjusting for market conditions. A 2.5x in 2010-2020 (zero rates, multiple expansion) is not equivalent to 2.5x in a higher-rate environment. Some LPs now risk-adjust using PME (Public Market Equivalent).

What sophisticated LPs do:

• Stress-test unrealized marks (what if portfolio is worth 50% less?)
• Haircut continuation fund "exits"
• Focus on DPI for mature funds (>7 years)
• Demand "TVPI without subscription lines" reporting
• Cross-reference with IRR to catch time-value distortions

The emerging consensus: For fund-of-funds or secondaries, blended TVPI requires weighted aggregation by commitment or NAV. Always pair TVPI with IRR and DPI for a complete picture. The metric isn't broken — but it requires context that standard quarterly reports don't provide.