Asia Private Equity Institute (APEI) Private Equity Insights Q3 212 Contents An Introduction to the APEI The Geography of Private Equity by Melvyn Teo Update on the Institute s Activities An Introduction to the APEI The vision of the Asia Private Equity Institute is to be the premier research and knowledge hub for private equity and venture capital activities in the Asia Pacific region. According to Preqin, Asia PE funds raised $62 billion in 211, double the $31 billion raised in 21. This 1% growth rate was approximately twice that of the global PE fund raising market, and Asia funds have now taken over Europe as the 2 nd largest fund-raising market after North America. Despite the surge in interest for private equity in Asia, most academic and practitioner research on private equity remains focused on US and Europe. As the first Asia-focused academic research centre on private equity and venture capital, APEI seeks to fill this knowledge gap. To do so, APEI will be an integrated platform that (i) conducts high quality academic and applied research on private equity and venture capital (ii) educates practitioners and disseminates new ideas and methods, thereby raising the standards and level of professionalism in the industry (iii) elevates the profile of the private equity and venture capital industry in Singapore and Asia. Some of the activities of the institute include a quarterly private equity insights newsletter that parlays academic findings into key lessons for general partners and limited partners, a closed door quarterly investment roundtable which facilitates an exchange of investment ideas between key industry players in an intimate setting, and an annual private equity conference where academics, general partners, limited partners, and industry experts discuss and debate topical issues that resonate with the industry. 1
The Geography of Private Equity Melvyn Teo 1 Executive summary We explore the impact of geographic proximity on the performance of private equity funds. Our analysis focuses on general partners that invest outside the developed world. We find that nearby funds outperform distant funds by 8.49 percent per annum (IRR). The outperformance of nearby funds is pervasive across most investment regions, fund types, and vintage years, and also manifest in investment multiples and public market equivalents. Fund size, industry specialization, and geographical specialization heighten the effects of a local informational advantage. Distant funds are able to compensate for their geographical disadvantage by leveraging on financial, technical, and management expertise. On balance, the findings suggest that geographically proximate general partners enhance performance by leveraging on local information networks to source for and exit from deals. To achieve stellar returns, private equity firms need to source for the best deals, monitor their investments closely, provide savvy financial, technical, and management advice to their portfolio companies, and exit from deals in a timely fashion. Geographical proximity may help this process. Fund managers plugged into the local network may be better placed to tap into the most lucrative deals. Nearby funds may be better positioned to closely watch over their investment companies and give constructive advice that augments firm value. Funds that are in tune with the local market sentiment and connected to local players may be able to orchestrate a more effective exit strategy. Indeed, the close relationship that private equity firms have with their portfolio companies suggests that the impact of geographic proximity on investment performance should be greater than for hedge funds and mutual funds. We explore the impact of distance on private equity fund performance in this inaugural issue of the Private Equity Insight. We ask: Do nearby funds outperform distant funds? If so, how pervasive is the impact of geography? Does fund size, specialization, chronology, or partner expertise amplify or attenuate the effects of distance? Our analysis leverages on the Preqin private equity performance and cash flow databases. The performance database includes information on fund IRRs and multiples of invested capital while the cash flow database includes monthly capital calls and distributions information for a subset of funds in the performance database. The Preqin database also offers information on fund characteristics such as investment region, vintage, investment type, office location, industry specialization, partner expertise, etc. Preqin obtains its data primarily from public filings by pension funds, from Freedom of Information Act (FOIA) requests to public pension funds, and also voluntarily from GPs and LPs. Therefore, the fund sample is not entirely free of self-selection bias. Nonetheless, Harris, Jenkinson, and Kaplan (211) argue that the consistency of returns from three distinct datasets: Burgiss, Preqin, and Cambridge Associates, despite very different sample selection criteria, suggests that they are likely to represent reliable measures of private equity performance. Moreover, most of our analysis will focus on differences in the cross-section of fund performance. Therefore, our results are less 1 Melvyn Teo is Professor of Finance and Co-Director, Asia Private Equity Institute (APEI) at the Singapore Management University. E-mail: melvynteo@smu.edu.sg. Phone: +65-6828-735. 2
affected by selection issues unless that there are systematic differences in sample selection across the fund groups that we focus on, e.g., nearby versus distant funds. Most funds invested in the US are also located in the US. The same can be said for funds invested in Western Europe. Therefore, our study focuses on funds that invest outside the developed world, specifically Asia, Latin America, Africa, and the Middle East. In the Preqin sample, there are 373 such funds with performance data as of March 212. Panel A of Table 1 reports the average size, IRR, investment multiple, and public market equivalent (PME) for these funds. PME is calculated using the method of Kaplan and Schoar (25). First, all cash flows from (distributions) and to (capital calls) the fund are discounted using the total realized return of the MSCI Emerging Markets Index from the fund s inception to the distribution date as the discount rate. 2 The discounted outflows and inflows are then summed to obtain the total discounted outflows and total discounted inflows to the fund. PME is simply the ratio of total discounted outflows to the total discounted inflows and reflects the after fee return to private equity investments relative to public equities. 3 As shown in Table 1, the average fund in our sample delivers an IRR of 12.47 per year, generates cash flows that are 1.58 times that of invested capital, and outperforms public equities by about 1 percent. Table 1: Summary statistics Number of funds Average AUM Average IRR Average Multiple Average PME Panel A: All funds All funds 373 456.83 12.47 1.58 1.1 Panel B: Funds grouped by investment region Asia 245 57.54 12.53 1.64 1.17 Latin America 62 331.2 12.4 1.48.99 Africa 23 157.88 11.35 1.42 1.15 Middle East 43 145.4 13.33 1.52.84 Panel C: Funds grouped by vintage year pre 1996 44 18.48 13.71 2.27.77 1996-2 67 33.37 11.63 1.69.99 21-25 92 368.18 19.75 1.87 1.11 26-21 17 629.73 8.53 1.19 1.16 Panel D: Funds grouped by investment type Buyout 87 6.57 11.27 1.52 1.14 Growth 56 358.18 17.46 1.54 1.3 Real estate 62 914.79 8.35 1.16.96 Infrastructure 13 619.96 7.92 1.5.69 Venture 117 148.21 13.18 1.85 1.1 Others 38 38.4 13.92 1.71 1.1 2 Kaplan and Schoar (25) use the S&P 5 return for the calculation of PME as they investigate US focused funds. 3 Ending NAVs are treated as true values as in Kaplan and Schoar (25). 3
Panels B, C, and D report summary statistics for funds grouped by investment region, vintage year, and investment type, respectively. Asia focused funds comprise the largest group of funds in our sample followed by Latin America focused funds. Many funds were launched between 26 and 21, reflecting the recent growth in private equity in these markets. Funds with vintages between 26 and 21 have also raised significantly more capital than funds from earlier vintages. However, the average performance of the later vintages has trailed that of the earlier vintages at least when we look at IRR. The fund sample features a diversified pool of GPs including venture, buyout, real estate, growth, and infrastructure funds. While there are more venture capital funds than funds operating other investment strategies, venture capital funds manage the least capital on a per fund basis, i.e., US$148m. By contrast, the average real estate fund manages significantly greater capital, i.e., US$914m. Finally, growth funds have outperformed other investment types while infrastructure funds have underperformed. To get a basic understanding of the impact of geography on private equity performance, we split the fund sample into nearby and distant funds. Nearby funds are funds that have an office (headquarters or branch office) located within their investment region while distant funds are funds that have no offices located within their investment region. 4 This dichotomy yields 36 nearby funds and 67 distant funds. Nearby funds appear to suffer from a capital raising disadvantage relative to distant funds. They raise on average US$369m which is about US$5m less than the US$862 raised by distant funds. This finding echoes that of Teo (29) who show that distant hedge funds based in the US and UK, despite underperforming nearby funds, are able to raise more capital, charge higher fees, and set more onerous redemption terms than nearby hedge funds. Next we compare the IRRs, investment multiples, and PMEs of nearby versus distant funds. The results in Figures 1, 2, and 3 reveal that based on any performance metric, nearby funds significantly outperform distant funds. The outperformance based on average IRR is an impressive 8.49 percent per year (t-statistic = 2.93). This dominates the analogous spread between nearby and distant hedge funds of 6.31 percent per year in Teo (29). Figure 2 reveals that nearby funds deliver an investment multiple that is.39 more than that delivered by distant funds (t-statistic = 2.42) while Figure 3 shows that relative to public equities, nearby funds outperform by 29 percent more than do distant funds (t-statistic = 2.52). Clearly, geographical proximity is helpful for private equity performance. Figure 1: IRRs of nearby versus distant funds 14 12 nearby funds distant funds 1 IRR(%) 8 6 4 2 Average IRR Median IRR Weighted Average IRR 4 Our results weaken but remain qualitatively similar when we classify nearby and distant funds based only on fund headquarter location. 4
Figure 2: Multiples of nearby versus distant funds 1.8 1.6 nearby funds distant funds 1.4 Investment multiple 1.2 1.8.6.4.2 Average Multiple Median Multiple Weighted Average Multiple Figure 3: PMEs of nearby versus distant funds 1.4 1.2 nearby funds distant funds 1.8 PME.6.4.2 Average PME Median PME Weighted Average PME How pervasive are our results? In Figures 4,5 and 6, we sort the sample based on investment region, vintage year, and investment type, respectively, and graph the performance (IRR) of nearby and distant funds. The size of the bubbles indicate the average assets under management of the funds within each group. It is clear from Figure 4 that nearby funds outperform distant funds in all regions except Africa; the bubbles for all investment regions save Africa are all comfortably to the right of the dotted 45 degree line, which indicates that relative to distant funds, nearby funds deliver higher IRRs. When funds are sorted based on vintage year as in Figure 5, we see again that the majority of the bubbles are comfortably to the right of the 45 degree line. Only for vintages between 21 and 25 do we find that the outperformance of nearby funds is marginal. Figure 6 sorts funds by investment type. Clearly, nearby funds dominate their distant competitors for all investment styles except venture. 5 5 We show in Table 3 that the underperformance of nearby venture funds relative to distant venture funds may have to do with their smaller size. 5
Figure 4: Funds sorted by investment region 2 18 16 Average IRR of distant funds 14 12 1 8 6 africa asia 4 2 middle east & israel americas 2 4 6 8 1 12 14 16 18 2 Average IRR of nearby funds Figure 5: Funds sorted by vintage year 22 2 18 21-25 Average IRR of distant funds 16 14 12 1 8 6 4 2 26-21 1996-2 pre1996 2 4 6 8 1 12 14 16 18 2 22 Average IRR of nearby funds 6
Figure 6: Funds sorted by investment type 2 venture Average IRR of distant funds 15 1 5 infrastructure real estate others growth -5 buyout -1-1 -5 5 1 15 2 Average IRR of nearby funds One concern is that nearby funds may be smaller than distant funds. Since it is well known that capacity issues impinge on the performance of private equity funds, the performance spread between nearby and distant funds may simply be a by-product of the latter s sizeable assets under management. To test, we estimate crosssectional regressions on fund IRR, investment multiple, and PME. Both univariate and multivariate regressions which control for the logarithm of fund size and fixed effects for investment region and vintage year are estimated. The results displayed in Table 2 reveal that fund size has a detrimental effect on fund performance as measured by IRR and investment multiple. However, even after controlling for the negative impact of fund size on performance, nearby funds still outperform distant funds. The coefficient estimate on the LOCAL variable that takes a value of one when a fund has its headquarters or branch office located in its investment region, and a value of zero otherwise, displayed in the second column of Table 2, indicates that nearby funds have an IRR of 7.15 percent greater than that of distant funds after controlling for other factors. This result is statistically significant at the five percent level. Inferences do not change when we analyze investment multiples and PMEs. Table 2: Regressions on fund location Dependent variables Independent variables IRR IRR Multiple Multiple PME PME LOCAL 8.96** 7.15**.43**.31**.29**.37* (2.96) (2.47) (3.74) (3.9) (3.18) (2.8) Log(AUM) -3.91** -.15** -.1 (-4.9) (-3.26) (-.38) Region Fixed Effects No yes No Yes no yes Vintage Fixed Effects No yes No Yes no yes 7
Number of observations 36 36 313 313 96 96 Note: *significant at the 5% level; **significant at the 1% level Next, we dig deeper and explore the factors that sharpen or attenuate the informational advantage held by nearby general partners. In Table 3, we report the impact of fund size, specialization, and chronology on the performance spread between nearby and distant funds. Panel A divides the sample equally into small, medium, and large funds based on capital raised. The monotonic increase in the nearby minus distant fund IRR spread as we move from the small fund tercile to the large fund tercile suggests that geographic proximity partly ameliorates the deleterious effect of fund size on private equity performance. When we sort funds by geographic specialization in Panel B and by industry specialization in Panel C, we find that specialization amplifies the benefits of local information. Nearby funds outperform distant funds even more for funds that invest in a single country or focus on a few industries than for funds that invest in multiple regions or have footprints in many industries. Collectively, these results indicate that close and deeper information networks are more helpful for investment performance than are wide and more diffuse information networks. A natural corollary is that the opportunities to take advantage of local information are limited. In line with this view, Panel D indicates that first time funds benefit more from local information than do subsequent follow-on funds raised by the same private equity firm. Table 3: Funds sorted by size, specialization, and chronology Number of funds Average AUM (US$m) Average IRR (%) Nearby funds Distant funds Nearby funds Distant funds Nearby funds Distant funds Nearby-Distant Panel A: Funds grouped by fund size Small funds 17 12 56.5 5.88 18.45 18.22.24 Medium funds 16 13 229.4 298.61 14.15 12.18 1.97 Large funds 83 39 951.2 1299.8 9.21 -.86 1.7** Panel B: Funds grouped by geographic specialization Single country 132 15 193.62 537.7 16.5 4.57 11.93 Single region 158 41 58.55 664.64 11.94 3.31 8.63** Multiple regions 16 11 429.66 1975.9 13.53 14.96-1.43 Panel C: Funds grouped by industry specialization Few industries 131 53 291.74 14.8 12.68 3.85 8.83** Many industries 157 8 448.91 147.1 15. 16.8-1.8 Panel D: Funds grouped by fund chronology First time funds 117 23 264.55 388.38 16.82 4.35 12.46** Follow on funds 184 41 44.25 1194.4 11.17 4.17 6.99* Note: *significant at the 5% level; **significant at the 1% level 8
As discussed, geographical proximity may help private equity firms in a variety of ways. Nearby funds may be better placed to tap into the most attractive deals, monitor their investments, give better technical, operational, and management advice that augments firm value, and exit from deals in a timely fashion. An investigation into the effects of firm expertise on the informational advantage of nearby funds may shed light on the exact mechanism by which geographical proximity confers a performance advantage for nearby funds. In that effort, we test whether the financial, technical, management, and general expertise of private equity firms impacts the benefit conferred by geographical proximity. Table 4 reports the nearby minus distant fund IRR spread for funds segregated by expertise. The results broadly indicate that expertise ameliorates the impact of the local informational advantage. For example, within the group of funds with no financial expertise, nearby funds outperform distant funds by 9.43 percent per year. Conversely, within the group of funds with financial expertise, nearby funds outperform distant funds by only 6.54 percent. The differences are starker when we sort funds by technical and industry expertise. For instance, within the group of funds without industry expertise, nearby funds outperform distant funds by an impressive 1 percent per year while within the group of funds with industry expertise, the analogous spread is only 2.73 percent. On balance, this suggests that general partner expertise can compensate for distance and that the outperformance amongst nearby funds has less to do with frequent monitoring and giving better advice, and more to do with superior deal sourcing and timely exits. Table 4: Funds sorted by expertise Number of funds Average AUM (US$m) Average IRR (%) Nearby funds Distant funds Nearby funds Distant funds Nearby funds Distant funds Nearby-Distant Panel A: Funds grouped by financial expertise No financial expertise 166 59 321.51 872.8 14.8 5.37 9.43** Financial expertise 14 8 425.24 782.13 13.4 6.5 6.54 Panel B: Funds grouped by technical expertise No technical expertise 276 66 391.61 869.5 14.3 5.5 8.53** Technical expertise 3 1 154.34 45.22 13.67 5.9 7.77 No industry expertise 184 61 285.27 859.8 15.1 5.1 1.** Industry expertise 122 6 49.51 892.82 12.32 9.58 2.73 No operational expertise 161 6 312.8 868.4 14.12 4.8 9.32** Operational expertise 145 7 431.92 86.19 13.844 11.543 2.3 Panel C: Funds grouped by management expertise No mgt expertise 197 62 331.56 846.22 13.31 5.29 8.1** Mgt expertise 19 5 434.63 151.4 15.23 8.16 7.7 No strategic expertise 149 61 326.36 859.8 13.23 5.1 8.13** Strategic expertise 157 6 47.7 892.82 14.71 9.58 5.13 Panel D: Funds grouped by general expertise No recruiting expertise 231 63 349.86 854.91 15.43 5.44 9.99** Recruiting expertise 75 4 426.6 972.23 9.57 6.55 3.2 No networking expertise 149 59 31.9 872.8 15.45 5.37 1.8** Networking expertise 157 8 431.94 782.13 12.61 6.5 6.11 Note: *significant at the 5% level; **significant at the 1% level 9
Conclusion In this issue of the Private Equity Insight, we have marshaled significant evidence to suggest that geographic proximity benefits private equity fund performance. Our findings provide valuable lessons for players in the private equity space about the nature of the local informational advantage. First, we show that geographic proximity is particularly helpful for funds with significant assets under management. Second, closer and deeper information networks are more helpful for investment performance than are wider and more diffuse information networks. Therefore, private equity firms that operate in a single country or in a few industries benefit most from geographic proximity. Third, distant private equity firms can leverage on their financial, technical, and management expertise to overcome some of the geographical disadvantage. This suggests that the stellar performance of nearby funds has less to do with their ability to monitor and advise their portfolio companies, and more to do with their ability to source for and exit from deals. Finally, it is unlikely that the local informational advantage will be eroded away quickly by an influx of nearby funds given the capital raising advantage that distant firms in developed countries possess. References Harris, R., Jenkinson, T., Kaplan, S., 211. Private equity performance: what do we know? Unpublished working paper, University of Virginia. Kaplan, S., Schoar, A., 25. Private equity performance: returns, persistence, and capital flows. Journal of Finance 6, 1791-1823. Teo, M., 29. Geography of hedge funds. Review of Financial Studies 22, 3531-3561. 1
Update on the Institute s Activities Education The APEI organized our first annual private equity conference on Sept 11 th 212. Tim Dattels, a general partner from TPG, and Tim Jenkinson, an academic from Oxford University, kicked off the conference and delivered exceptional keynote speeches on the opportunities in Emerging Markets and on the performance of private equity relative to public equity, respectively. The institute co-director, Melvyn Teo, then took the opportunity to present his work on the Geography of Private Equity which resonated with the mainly Asiabased practitioner audience. Finally, we rounded up the conference with a limited partner panel which was moderated by Lily Fang from INSEAD. The panel consisting of a trio of limited partners including Ashok Samuel from GIC, Marc Lau from Axiom Asia, and Sunil Mishra from Adams Street discussed the theme: Do the returns of private equity compensate for the risks? The panel also touched upon issues such as the challenges related to fund size, the public perception of the private equity industry across the world, and how private equity adds value to companies and impacts job creation. The conference was well attended with 164 attendees, mostly practitioners, including institutional investors, family offices, and private wealth. For more information regarding the Asia Private Equity Institute (APEI) at SMU and our upcoming activities, please contact Ms Karyn Tai, centre coordinator (Tel: +65-6828-933, E-mail: apei@smu.edu.sg). We look forward to receiving your suggestions and comments. 11