The Fund Metrics That Actually Matter: TVPI, DPI, IRR — and How to Track Them Live

Why Fund Metrics Matter More Than Ever

If you’re managing a venture capital or private equity fund, you’re drowning in data. Portfolio company valuations shift monthly. Cash flows are unpredictable. LPs demand quarterly updates. And somewhere in the noise, you need to know: Is this fund actually performing?

That’s where fund metrics come in. But not all metrics are created equal. Many GPs spend weeks assembling spreadsheets to calculate TVPI, DPI, and IRR—three numbers that, taken together, tell the complete story of fund performance. The problem is that these calculations happen in static Excel files, refreshed quarterly or annually. By the time you know your true metrics, market conditions have shifted.

This guide walks you through the fund metrics that matter, why they matter, and how to build live dashboards in Apache Superset so you can track fund performance in real time. We’ll cover the formulas, the benchmarks, the pitfalls, and the implementation patterns that work at scale.

Understanding the Core Three: TVPI, DPI, and IRR

Let’s start with definitions, because precision matters here.

Total Value to Paid-In Capital (TVPI), often called the “multiple,” is the simplest metric: it’s the ratio of a fund’s total value (distributions plus remaining portfolio value) divided by the capital invested so far. If a fund has paid in $100M and the portfolio is worth $250M total, the TVPI is 2.5x. That’s a clean, intuitive number that LPs understand immediately.

Distributions to Paid-In Capital (DPI) is the cash you’ve actually returned to LPs. It’s distributions divided by paid-in capital. If you’ve returned $150M on $100M invested, your DPI is 1.5x. This is the only metric that reflects real money back in LP pockets. Everything else is paper gains.

Internal Rate of Return (IRR) is the annualized return accounting for the timing and size of cash flows. It’s the discount rate that makes the net present value of all cash flows equal to zero. IRR is harder to calculate than TVPI or DPI, but it’s the metric that lets you compare a fund’s performance to benchmarks, other funds, and public market returns. A 15% IRR over 10 years is strong; 25% is exceptional.

These three metrics are not interchangeable. They tell different stories:

• TVPI tells you total value creation (realized and unrealized).
• DPI tells you cash returned (the only metric that matters until you exit).
• IRR tells you the rate at which capital compounded, adjusted for timing.

A fund can have high TVPI but low DPI if most value is still locked in portfolio companies. A fund can have high DPI but low IRR if returns came late in the fund lifecycle. Understanding all three is essential.

The Math: Formulas You Need to Know

Let’s get concrete. The formulas are straightforward, but the data quality behind them is everything.

TVPI Formula

TVPI = (Cumulative Distributions + Remaining Portfolio Value) / Cumulative Paid-In Capital

Example: A fund has paid in $100M. Current distributions are $60M. Remaining portfolio value is $180M.

TVPI = ($60M + $180M) / $100M = 2.4x

This fund has created $240M in total value on a $100M investment.

DPI Formula

DPI = Cumulative Distributions / Cumulative Paid-In Capital

Using the same fund:

DPI = $60M / $100M = 0.6x

Only 60% of the capital has been returned in cash. The rest is still in the portfolio.

IRR Formula

IRR is more complex. It’s the rate (r) that solves:

0 = -CF₀ + CF₁/(1+r)¹ + CF₂/(1+r)² + ... + CFₙ/(1+r)ⁿ

Where CF is each cash flow (negative for capital called, positive for distributions and final residual value), and the subscript is the year.

You won’t calculate this by hand. Use a spreadsheet XIRR function or a dedicated analytics platform. But understanding the concept is critical: IRR accounts for when money flows in and out. A fund that returns 2x in year 5 has a different IRR than one that returns 2x in year 10.

For understanding key venture capital metrics including IRR, MOIC, and gross IRR, the formulas and timing assumptions become clearer when you see them applied across fund vintages.

Gross vs. Net: The Metric That Changes Everything

Here’s where many GPs trip up: there are two versions of every metric—gross and net.

Gross metrics are calculated before fees and expenses. Net metrics are after. For TVPI and DPI, the difference is usually 5–15%, depending on fee structure. For IRR, the difference can be 2–4 percentage points annually.

LPs care about net. GPs often report gross (to look better). The best practice is to report both, clearly labeled.

Let’s say a fund’s gross TVPI is 2.5x, but after management fees, fund expenses, and carry, the net TVPI is 2.1x. That’s a real difference. And it’s the net number that determines whether LPs will commit to the next fund.

When you build dashboards, calculate both. Make it easy to toggle between them. And be explicit about what’s included in “expenses”—management fees, legal, audit, insurance, and platform costs all matter.

Benchmarking: What’s Good?

Metrics only mean something in context. Here’s what the data says:

For venture capital funds:

• A TVPI of 2.0x or higher is considered strong. Below 1.5x is concerning.
• A DPI of 0.5x at the midpoint of the fund lifecycle (year 5–6) suggests healthy distributions. Below 0.3x means you’re still waiting for exits.
• An IRR of 15% or higher is competitive. 20%+ is exceptional. Below 10% underperforms public markets.

For private equity funds:

• TVPI expectations are higher: 2.5x–3.5x for buyout funds.
• DPI is usually higher too, because PE funds focus on cash distributions.
• IRR targets are often 20%+, depending on the vintage and strategy.

These benchmarks vary by stage (seed vs. growth), geography, and sector. But they give you a baseline. If your fund is at year 7 with a TVPI of 1.2x and DPI of 0.2x, you’re behind. If it’s 2.8x and 1.1x, you’re ahead.

According to research on evaluating venture capital performance metrics, these benchmarks shift based on fund stage and LP expectations, making real-time tracking essential for course correction.

The Hidden Metrics: RVPI, MOIC, and Why They Matter

TVPI, DPI, and IRR are the big three, but they don’t tell the whole story. Two other metrics provide critical context.

Residual Value to Paid-In Capital (RVPI) is the unrealized multiple. It’s remaining portfolio value divided by paid-in capital.

RVPI = Remaining Portfolio Value / Cumulative Paid-In Capital

RVPI + DPI = TVPI. So if TVPI is 2.4x and DPI is 0.6x, then RVPI is 1.8x. This tells you how much value is still on the table. A high RVPI fund is illiquid; a low RVPI fund is mostly cashed out.

Multiple on Invested Capital (MOIC) is sometimes used interchangeably with TVPI, but technically it’s the total value multiple without the time component. For most purposes, MOIC and TVPI are the same thing.

These metrics matter because they tell you about portfolio composition and exit timing. A mature fund should have low RVPI and high DPI. An early-stage fund should have high RVPI and low DPI.

For detailed guidance on fund performance metrics including TVPI, DPI, and RVPI, understanding how these metrics interconnect reveals portfolio health at a glance.

Building Live Fund Dashboards: The Implementation

Now for the practical part: how do you actually track these metrics in real time?

Most GPs start with Excel. You have a cap table spreadsheet, a portfolio valuation tracker, and a cash flow log. Every quarter, someone spends two weeks pulling data, calculating metrics, and emailing PDFs to LPs. It’s slow, error-prone, and disconnected from reality.

A better approach is to build a live dashboard in Apache Superset, a modern open-source BI platform that handles complex calculations, integrates with your data sources, and lets you query data with natural language.

Here’s the architecture:

Data Layer

Your source of truth should be a single database or data warehouse. This could be:

• A PostgreSQL database that syncs with your cap table software (like Carta or Pulley).
• A data warehouse (Snowflake, BigQuery, Redshift) that aggregates cap table, portfolio valuation, and cash flow data.
• An API layer that pulls from multiple sources and lands data in a warehouse.

The key is that all three data streams—paid-in capital, distributions, and portfolio value—come from a single source and update regularly (daily or weekly).

Metric Calculation Layer

In Superset, you’ll create SQL-based calculated columns or use Superset’s native aggregation features to compute:

• Cumulative paid-in capital (sum of all capital calls).
• Cumulative distributions (sum of all cash returned).
• Remaining portfolio value (sum of current valuations of all holdings).
• TVPI, DPI, RVPI, and IRR derived from the above.

For IRR, you’ll likely need a stored procedure or a Python UDF, since IRR requires iterative calculation. D23’s managed Apache Superset platform supports custom Python code and SQL functions, making IRR calculation straightforward.

Dashboard Layer

Your dashboard should show:

1. Fund-level summary cards: TVPI, DPI, IRR, and RVPI at a glance, with gross and net versions.
2. Time series charts: How TVPI, DPI, and IRR have evolved over the fund’s life. This shows whether you’re on track.
3. Portfolio composition: Holdings by stage, sector, and geography, with realized and unrealized gains.
4. Cash flow waterfall: Paid-in capital, distributions, and remaining value stacked over time.
5. Benchmark comparison: Your metrics vs. industry benchmarks by vintage and stage.
6. Company-level drill-down: Click into any holding to see its contribution to TVPI, DPI, and IRR.

The beauty of a live dashboard is that you can update portfolio valuations weekly, and your metrics update automatically. No manual calculation. No stale data.

Handling Valuation: The Data Quality Challenge

Here’s where most fund dashboards break down: valuation data is messy.

Portfolio companies are valued on different schedules. Some use recent financing rounds. Others use revenue multiples. Some are marked down. Some are marked up without clear justification. And valuations change constantly as companies grow, pivot, or struggle.

When you build a live dashboard, you need a clear valuation policy:

• Financing round: If a company has raised recently, use that valuation (typically the post-money valuation of the most recent round).
• Revenue multiple: For mature companies, use a revenue multiple (e.g., 5x ARR) and update it quarterly.
• Comparable transactions: If comparable companies have been acquired, use those prices.
• Discounted cash flow: For later-stage companies, DCF models can inform valuation.
• Write-down: If a company is struggling, be conservative. Write it down.

The key is consistency. Use the same methodology every quarter. Document assumptions. And be conservative—it’s better to surprise LPs with upside than disappoint them with write-downs.

In your Superset dashboard, create a valuation audit trail. Show the methodology used for each holding, the valuation date, and any changes quarter-over-quarter. This transparency builds LP trust.

Text-to-SQL and AI for Fund Analytics

One of the most powerful features in modern BI platforms is text-to-SQL—the ability to ask questions in natural language and get answers instantly.

Imagine asking your dashboard: “Show me the TVPI of my Series A companies that have raised in the last 18 months.” Or “Which companies have the highest unrealized gains?” Or “What’s our IRR if we exit all companies at current valuations?”

With D23’s AI-powered analytics capabilities, you can ask these questions and get instant answers. The system translates your question into SQL, queries the database, and returns visualizations.

For fund analytics, this is transformative. Instead of waiting for your data team to run reports, you can explore your portfolio in real time. You can test scenarios (“What if we exit this company at 3x?”). You can spot trends (“Are our Series B companies growing faster than last year?”).

This is especially valuable for LPs. Instead of sending them a static PDF, you can give them access to a live dashboard where they can ask their own questions. This builds confidence and reduces support burden.

API-First Fund Dashboards: Embedding and Sharing

Many GPs need to share fund metrics with LPs, advisors, and board members. But not everyone should have access to your full database.

Superset’s API-first architecture lets you:

1. Embed dashboards in your GP website or LP portal. LPs see real-time metrics without logging into Superset.
2. Generate reports programmatically. Every Friday, a script exports the week’s metrics and emails them to stakeholders.
3. Build custom views for different audiences. LPs see TVPI, DPI, and IRR. Your investment team sees company-level details. Your finance team sees fee calculations.
4. Control access with role-based permissions. You decide who sees what.

This is where D23’s API-first approach to embedded analytics shines. You can build a complete LP portal around Superset, with dashboards, reports, and data downloads—all without building custom software.

Real-World Example: A Series A Fund Dashboard

Let’s walk through a concrete example. You’re managing a $200M Series A fund, now in year 4 of a 10-year lifecycle.

Current state:

• Paid-in capital: $185M (92% of fund size)
• Cumulative distributions: $45M (mostly from one IPO exit)
• Remaining portfolio value: $320M (based on recent rounds and benchmarking)
• Gross TVPI: 1.97x
• Net TVPI (after fees and carry): 1.65x
• Gross DPI: 0.24x
• Net DPI: 0.20x
• Gross IRR: 22% (annualized)
• Net IRR: 18%

What this tells you:

You’re ahead of benchmark for a year-4 Series A fund. Your TVPI is strong. But your DPI is low—you’ve only returned 20% of capital. This is normal for Series A, which has longer holding periods. Your IRR is solid, suggesting that the one exit was well-timed and the remaining portfolio is valuable.

Dashboard breakdown:

1. Fund summary card: Shows all six metrics (gross and net TVPI, DPI, IRR) with year-over-year changes.
2. TVPI waterfall: Shows how TVPI has grown from 1.2x (year 1) to 1.97x (year 4), with contributions from each exit and valuation increases.
3. Portfolio composition: 45 companies, color-coded by stage and geography. Size of each bubble represents portfolio value.
4. Cash flow timeline: Shows capital calls, distributions, and remaining value over the fund’s life.
5. Benchmark comparison: Your metrics vs. median Series A fund by vintage. You’re in the 75th percentile for TVPI, 50th for DPI (expected).
6. Company drill-down: Click any company to see its valuation history, cost basis, current value, and contribution to fund metrics.

This dashboard updates weekly. When a portfolio company raises a new round, valuation updates automatically. When you make a distribution, DPI updates. When you write down a struggling company, TVPI adjusts. Your LPs can log in and see current metrics anytime.

Common Pitfalls and How to Avoid Them

Building fund dashboards is straightforward in theory, but implementation reveals tricky edge cases.

Pitfall 1: Inconsistent valuation methodology

You mark one company at 5x revenue, another at 3x. When valuations change, you’re not sure if it’s real growth or methodology drift.

Solution: Document your valuation policy. Use the same methodology for all companies in a cohort. Review and adjust annually, but stay consistent quarter-to-quarter.

Pitfall 2: Forgetting to include fees in net metrics

You calculate net TVPI but forget to subtract management fees, carry, or expenses. Your net metrics are still too high.

Solution: Create a separate fee schedule in your database. Calculate fees as a time-based accrual (management fee) and a distribution-based accrual (carry). Subtract both from net metrics.

Pitfall 3: IRR calculation errors

You use a simple XIRR function, but it doesn’t account for partial exits, write-downs, or follow-on investments. Your IRR is wrong.

Solution: Use a proper IRR calculation that models each cash flow event. Test against benchmarks. And remember: IRR is sensitive to timing. A distribution in December vs. January can change IRR by 1–2 percentage points.

Pitfall 4: Stale data

Your dashboard shows data from last month. By the time you see a problem, it’s too late to act.

Solution: Automate data ingestion. Sync with your cap table software daily or weekly. Build alerts for significant changes (valuation drops, unexpected distributions). Make the dashboard a real-time tool, not a quarterly report.

Advanced Patterns: Scenario Analysis and Forecasting

Once you have live metrics, the next step is scenario analysis. What if you exit company X at 4x? What if company Y gets acquired at its current valuation? How does that change fund IRR?

Superset supports this through calculated columns and dynamic filters. You can build a “scenario” table where you model different exit prices, timing, and valuations. Then compare scenarios side-by-side.

For example:

• Base case: Current valuations, expected exit timing. Projected fund IRR: 20%.
• Bull case: All companies exit at 1.5x current valuation, on schedule. Projected IRR: 28%.
• Bear case: 30% of portfolio written down, exits delayed 2 years. Projected IRR: 12%.

You can share these scenarios with LPs to show the range of possible outcomes. And as the fund evolves, you update assumptions and recalculate.

For practical guidance on using IRR, DPI, TVPI, and MOIC across the fund lifecycle, scenario analysis becomes a critical tool for decision-making at each stage.

Integration with Your Data Stack

Most GPs use multiple tools: Carta or Pulley for cap tables, Forge or similar for valuations, Stripe or custom systems for distributions. Getting all this data into one place is the hard part.

Superset integrates with:

• SQL databases: PostgreSQL, MySQL, Snowflake, BigQuery, Redshift. If your data is in a warehouse, Superset can query it.
• APIs: You can build a data pipeline that pulls from Carta’s API, aggregates with other sources, and lands in a warehouse. Superset queries the result.
• CSV uploads: For one-off data, you can upload CSVs directly into Superset.

The best approach is a data warehouse (Snowflake or BigQuery) that syncs with your cap table tool via an API integration (Fivetran or similar). Then Superset queries the warehouse. This gives you a single source of truth.

For comprehensive guidance on VC fund metrics and their interconnection, understanding how to structure your data layer ensures metrics are reliable and auditable.

Reporting and Compliance

LPs expect transparency. You need to document:

• How metrics are calculated.
• What data sources are used.
• When data is refreshed.
• Any assumptions or adjustments.

Your dashboard should include a “methodology” section that explains all of this. And you should audit your metrics quarterly against your GP records.

For regulatory compliance (especially if you’re a registered investment adviser), you may need to file reports with the SEC. TVPI, DPI, and IRR are standard metrics in these filings. Having a live, auditable dashboard makes compliance easier.

Getting Started: The MVP Approach

You don’t need a perfect system from day one. Start with an MVP (minimum viable product):

1. Month 1: Export your cap table and portfolio valuations into a spreadsheet. Calculate TVPI, DPI, and IRR manually. This is your baseline.
2. Month 2: Load this data into Superset. Create simple dashboards showing fund-level metrics and portfolio composition.
3. Month 3: Integrate with your cap table API. Automate data ingestion so metrics update weekly.
4. Month 4+: Add scenario analysis, company-level drill-downs, benchmark comparisons, and LP portal access.

Each step adds value. And because Superset is open-source and D23 provides managed hosting with expert support, you don’t need a large data engineering team to get started.

Why Superset for Fund Analytics

You might be wondering: why Superset instead of Looker, Tableau, or Power BI?

Three reasons:

1. Cost: Superset is open-source. You pay for hosting and support, not per-seat licensing. For a fund with dozens of LPs, this is a huge advantage.
2. Flexibility: Superset lets you write custom SQL, Python code, and complex calculated columns. Fund metrics often require custom logic that pre-built tools can’t handle.
3. API-first architecture: Superset was built for embedding and API access. You can embed dashboards in your LP portal, generate reports programmatically, and build custom applications on top.

For exploring VC metrics and their interpretation at different fund stages, a flexible platform that handles custom calculations and complex data relationships is essential.

Looker and Tableau are powerful, but they’re expensive and less flexible. Power BI is cheaper but less open. Superset strikes the right balance for fund analytics.

Conclusion: From Spreadsheets to Live Dashboards

Fund metrics—TVPI, DPI, and IRR—are the language of venture capital. They tell you whether you’re winning or losing, and they determine whether LPs will commit to your next fund.

But these metrics are only useful if they’re current. A dashboard built last quarter is already stale. By the time you see a problem, the market has moved.

The solution is a live dashboard in Apache Superset, updated daily or weekly, that shows real-time fund performance. This gives you visibility into your portfolio, confidence in your metrics, and a tool to share with LPs.

Starting is simple: load your cap table and portfolio data into Superset, calculate your core metrics, and build a basic dashboard. From there, you can add sophistication—scenario analysis, forecasting, benchmarking, and LP portal access.

The GPs who win are the ones who know their numbers. Make sure you’re one of them.

---

Appendix: Metric Formulas at a Glance

TVPI = (Cumulative Distributions + Remaining Portfolio Value) / Cumulative Paid-In Capital

DPI = Cumulative Distributions / Cumulative Paid-In Capital

RVPI = Remaining Portfolio Value / Cumulative Paid-In Capital

IRR = The discount rate that makes NPV of all cash flows equal to zero (use XIRR function in Excel or equivalent in your BI tool)

Net metrics = Gross metrics minus management fees, expenses, and carry

MOIC ≈ TVPI (technically, MOIC is the total value multiple; TVPI is MOIC adjusted for time)

For deeper understanding of how these metrics interconnect and drive fund performance assessment, refer to the external resources linked throughout this guide and experiment with your own fund data in a live dashboard.