For Institutions
For endowments, pensions and foundations
You cannot forecast what a venture allocation will return. You can forecast when it comes back. That asymmetry is what this instrument is built against.
Who speaks to investors
The program is in preparation and no securities are being offered through this site.
When it is ready, investor communications will be handled solely by a licensed investment adviser representative of a registered investment adviser. No one on this page solicits investors, discusses the terms of an investment, negotiates with investors, or accepts investor funds. Their work is with entrepreneurs and their companies.
No one on this page is paid a percentage of any financing, and no fee or participation is paid out of the structure to any broker or adviser.
The one thing you can forecast
Across 20,000 simulated funds, the dollar-weighted date on which capital comes back varies by a tenth. What comes back varies by a third.
| measure | coefficient of variation | 10th to 90th percentile |
|---|---|---|
| duration | 10.0 per cent | 8.19 to 10.63 years |
| DPI | 34.3 per cent | 1.02x to 2.51x |
| quartile | DPI | duration | return of the first 100 | DPI reaches 1.0x | 90% back by | net IRR |
|---|---|---|---|---|---|---|
| bottom | 1.03x | 8.96 | 8.53 | 11.1 | 11.7 | 0.56% |
| second | 1.49x | 9.29 | 8.12 | 9.9 | 12.1 | 6.66% |
| third | 1.86x | 9.54 | 7.85 | 9.3 | 12.4 | 10.34% |
| top | 2.52x | 9.80 | 7.57 | 8.6 | 12.8 | 15.18% |
| all | 1.73x | 9.40 | 7.97 | 9.6 | 12.2 | 9.03% |
9 per cent of simulated funds never return 1.0x inside 18 years.
What the model assumes
- 20,000 simulated funds, 30 companies each, a 2026 vintage, top-quartile access.
- 83 of every 100 committed reaches companies; the rest is management fee over the life.
- European waterfall, 20 per cent carry on proceeds above the commitment.
- 30 per cent of the middling names zombify into years 12 to 16: extensions, continuation vehicles, strip sales of what is left.
- The largest outcomes distribute stock over three years after a lockup.
- Figures are per 100 of LP commitment, net to the LP after fees and carry.
| outcome | probability | gross multiple | exit year |
|---|---|---|---|
| write-off | 40% | 0.00x | 3 to 7 |
| sub-1x | 20% | 0.10x to 0.90x | 4 to 9 |
| 1-3x | 22% | 1.00x to 3.00x | 5 to 10 |
| 3-10x | 13% | 3.00x to 10.00x | 6 to 12 |
| 10x+ | 5% | 10.00x to 30.00x | 7 to 13 |
These are modelled, not observed. The level assumptions are deliberately sober; the timing assumptions are the ones worth arguing with.
The tail is not only the failures
The correlation between a fund's DPI and its own duration is +0.34. The good funds are slower. The top quartile has the longest duration, 9.80 years, and does not have 90 per cent of its money back until year 12.8; the bottom quartile pays earliest because it pays so little. 23 per cent of all distributions are still outstanding at year 12, and that 23 per cent carries 1.03 years of the duration. It is a mix of the best asset in the fund and the worst.
11.40
If the exit window stays shut two more years, duration goes to 11.40 from 9.40.
Where the Note lands
| year | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| cumulative | 0.0 | −22.0 | −45.0 | −64.0 | −77.7 | −84.5 | −80.6 | −68.6 | −47.8 | −24.8 | −4.2 | 11.9 | 29.0 | 34.5 | 36.9 |
The deepest point is −84.5 per 100 committed, in year 5. The book is not self-funding until year 10.3. The Note's principal comes back in year five — 5.000 years of duration on the tax-exempt class, 4.867 on the taxable, where a small coupon pulls it forward. That is not a coincidence: both dates are set by the same deployment arithmetic.
5.000
the Note's duration, tax-exempt class
a single payment at year 5
4.867
the Note's duration, taxable class
coupon of 1.4291 a year, then 100.00
0
dispersion of that date
the same date in every outcome
Blending the two
| Note share | duration | sd of the date |
|---|---|---|
| 0 per cent | 9.40 | 0.94 |
| 20 per cent | 8.52 | 0.75 |
| 33 per cent | 7.93 | 0.63 |
| 50 per cent | 7.20 | 0.47 |
sd of the date is the standard deviation, across simulated funds, of the dollar-weighted return-of-capital date. The Note contributes none of it.
What this is not
It is a barbell, not a duration match.
Five years against 9.40 is roughly half. The term is capped at five years, so no version of this matches a venture book maturity for maturity. What it does is sit in the trough: the capital that lets an allocator hold the long duration through years five to nine without selling into the secondary market.
It floors the terminal value, not the mark.
The position marks as a five-year Treasury plus a warrant. A hundred basis points of rates moves the Treasury leg by roughly four and a half per cent, and the warrant marks with the company. What has a floor is what the position pays at maturity, not what it is carried at in between.
The floor is not a return.
If the company fails the unit pays 107.15 per 100. That is below every policy benchmark. The hurdle table says what company growth it takes to match each one:
| to match over five years | per 100 | one round | two rounds |
|---|---|---|---|
| U.S. Treasuries | 129.87 | 2.4% | 6.4% |
| policy portfolio at 6% | 133.82 | 5.3% | 9.5% |
| policy portfolio at 7% | 140.26 | 9.3% | 13.8% |
| policy portfolio at 8% | 146.93 | 13.0% | 17.7% |
One round / two rounds = how far the Warrant is diluted by later financings before it is exercised.
Against a venture sleeve the comparison runs the other way: a failed company returns roughly nothing and takes thirteen years to say so, where a failed Note returns 107.15 on a known date in year five. Failure in venture has a long duration. Failure here has a five-year one. What the structure does is convert selection risk into opportunity cost. It does nothing about selection itself.