Private equity investors are devoting increasing attention to return and its sources; the impact of management fees and carried interest on performance are under the spotlight. These concerns have led to renewed LP interest in co-investments, which allow LPs to provide additional equity finance to specific portfolio companies on the same terms as the host fund. There is typically no annual fee or carried interest payable on co-investments, which significantly reduces the cost. So if an LP can participate in attractive deals from a variety of GPs, it can build a high-return and low-cost portfolio.
About the study
Co-investment funds differ from primary private equity funds in that they group investments managed by different GPs. To explore the impact of manager diversification, we used proprietary data on 121 primary funds and their underlying portfolio companies to simulate 10,000,000 co-investment portfolios. Each simulated co-investment portfolio was based on a reference primary fund, and contained the same number of investments, but each investment was from a different manager. We also ensured the investments selected for the co-investment funds were from the same geography, roughly the same size and were invested within one year of the reference investment in the primary fund. We then computed aggregated gross multiples for both the primary and simulated funds. The differences between the distribution of multiples for the two types of funds reflect the impact of manager diversification.
The underlying company data (which includes 1,300+ investments) are largely comprised of buyouts acquired and exited between 1986 and 2011, plus and some late-stage venture and special situation investments. Geographically, almost two-thirds of the investments were US-based with the remainder split between Europe and the rest of the world.
In terms of performance, the median multiple was 1.4x, and the maximum was close to 60x. For the funds themselves level, the median aggregate multiple was 1.6x and the maximum was 13.5x. (see figure 1).
It’s worth pointing out that although the data employed in this study are representative of co-investment opportunities, not every investment was actually open to co-investors. However, since there is no conclusive evidence of ‘adverse selection’ (i.e. GPs offering co-investment opportunities for poor or risky investments), this should not introduce a performance bias, as long as we proceed on the basis that the co-investor LP has access to substantial deal flow. The distribution of returns for the simulated co-investment portfolios can be seen in figure 2.
Less risk, better performance
Risk in private equity is a complex subject and different parameters have been proposed to reflect the peculiarities of the asset class. But for the sake of simplicity purposes – and given that our focus was on downside risk – we looked at the percentage of portfolios returning less than the invested amount: 14.9% for the primary funds, and 11.3% for the simulated co-investment funds. In other words, the simulated co-investment funds exhibit less risk of not returning the full invested amount.
Furthermore, the difference would be wider if fees were considered, since co-investments attract lower fees. Because only realized investments were used to calculate fund multiples, fee data for the overall fund cannot be overlaid to arrive at net multiples. However, if we suppose an average company holding time of five years and a 2% management fee on invested capital and no carry, the proportion of primary funds that return less than 1x invested capital after fees would increase to 24.0%. This clearly indicates that co-investment portfolios benefit from risk reduction stemming from manager diversification.
The benefits are even more pronounced at the net return level. In addition to reducing risk, manager diversification also yields direct performance benefits in co-investment portfolios. The median multiple for the simulated co-investment funds was 1.87x – significantly higher than the median multiple for the primary funds of 1.64x.
Furthermore, close to 60% of all simulated co-investment funds performed better than the median primary fund. These results indicate that, on average, co-investment funds outperform primary funds. And this outperformance is even more pronounced on a net multiple basis: by applying the same 2% fees on invested capital for five years, the median net primary fund multiple drops to 1.45x – and nearly 70% of all co-investment portfolios have higher multiples.
Although the benefits of manager diversification in co-investment funds resemble the benefits of diversification across different managers in portfolios of primary funds, it is interesting to note that manager diversification in co-investment funds does not reduce the variance of the return distribution. Therefore, manager diversification does not reduce the potential for extraordinary returns. In fact, co-investment portfolios perform better than primary funds for all percentiles of return distribution – and these performance benefits are further accentuated for net multiples given the impact of fees on the performance of primary funds.
Figure 3 shows a side-by-side comparison of the return distributions of primary funds and simulated co-investment portfolios. The results show that, compared to primary funds, co-investment portfolio returns have a higher median and a fatter right tail. It also shows that the higher variance of co-investment fund returns doesn’t translate into higher risk – since the right-shift of the overall return distribution offsets the potential increase in downside. Moreover, the higher volatility is reflected in the higher upside potential of co-investment portfolios.
Co-investors also have an additional advantage over investors in primary funds: the ability to choose specific investments. LPs with access to quality deal flow and good execution capabilities have the potential to assemble very attractive portfolios. From our simulation, for instance, we concluded that a co-investor LP with top-quartile selection abilities would be able to assemble a portfolio that performs better than 80% of the primary funds in the sample.
In conclusion, then: the additional layer of diversification in co-investment portfolios was shown statistically to reduce risk and increase the median multiple – and the magnitude of these benefits is even larger after fees, given that co-investments are usually offered free of management fee and carry.
The benefits of diversification across vintage years and number of funds are well documented for primary fund investments. But while such diversification reduces downside risk, at the same time it also reduces the potential for outperformance.
In contrast, manager diversification in co-investment portfolios lowers downside risk – but does not reduce the upside potential. This combination renders manager diversification in co-investment portfolios a proverbial win-win.
The success of a co-investment strategy depends, to a great extent, on access to high-quality co-investment deal flow – which LPs cannot achieve without strong and enduring relationships with quality GPs. But those LPs with the necessary resources and relationships can create a very attractive portfolio of co-investments as a natural complement to a portfolio of primary funds.