Horizon IRRs provide a suitable metric to compare the performance of the private equity and venture capital asset classes for a defined period of time. In addition to holding IRR and multiples data for almost 7,800 funds, Preqin’s Performance Analyst online service also holds quarterly cash flow data for over 3,000 individual private equity and venture capital funds, using this data to calculate horizon IRRs for a range of timeframes. The calculation uses the net asset value at both the beginning and end of the period, as well as the capital calls and distributions in between. Preqin’s horizon IRRs are annualized, dollar-weighted and net of management and carried interest fees.
As illustrated in the chart above, venture capital funds have outperformed all other private equity strategies over the one-year period to March 2015, with a horizon IRR of 20.1%. In this one-year period, buyout achieved a horizon IRR of 16.7%, fund of funds 15.1%, all private equity 14.2%, mezzanine 12.4% and distressed private equity 6.2%, showing that distressed private equity vehicles have pulled down the overall private equity horizon IRR. The only private equity strategy to exceed the venture capital one-year horizon IRR is buyout, which posted a horizon IRR of 21% in the 10-year period to March 2015.
Over the three-year period, mezzanine vehicles posted the best horizon IRR (16.7%), followed closely by buyout (16.5%) and venture capital funds (15.3%). Over the five-year period, the horizon IRRs for each of the fund types are not too dissimilar, ranging from 12.4% to 16.8%. In contrast to the one-year period, venture capital funds were the worst performers over the 10-year period, with a horizon IRR of 5.9%; the lowest across all time periods and fund types examined, indicative of venture capital’s poor performance through the financial crisis.
With venture capital fund valuations rising, it will be interesting to see how the fund type performs in the coming years, and if it can return the full amount of unrealized capital to investors in distributions.