Foreword. Best wishes, Tarek Chouman

Transcription

1 Global Private Equity Performance Series An international comparison of major markets May 2018

2

3 Foreword Welcome to this special report on the risk and return characteristics of major private equity and venture capital markets globally, part of efront s Global Private Equity Performance Series. This study represents, to the best of our knowledge, the most comprehensive publicly available analysis of returns, risk and liquidity among private equity s main geographies. But this is not meant to be a definitive report. It is more an essay, an attempt to draw reasonable conclusions from historic data, using the most powerful data analytics available on the market, from efront Insight. Private markets do not offer precise or timely information, and the nature of closed-ended funds introduces other complexities. These challenges are opportunities for those willing to invest the time in truly understanding the idiosyncrasies and stages of evolution in different geographic markets. In analysing private markets, there are rarely any right answers, only good questions. efront Insight helps the world s leading institutional investors to ask them. I hope you find what follows illuminating and inspiring in your own search for the risk and return profile that suits your organisation. Best wishes, Tarek Chouman

4 Contents: Introduction Main Findings Performance analysis of the major PE markets All private equity funds Leverage buyout funds Venture capital funds Methodology...31 efront Annual Report 2018

5 efront Annual Report 2018

6 Introduction When analysing private equity and venture capital, forming a sophisticated understanding of the characteristics of specific geographic markets is pertinent for a number of reasons. Firstly, such funds deal with businesses on-the-ground, usually of a modest size, and often exposed to domestic conditions. Risk and return will be governed by a host of local considerations, such as the ease of doing business and labour laws, through to more nebulous factors such as entrepreneurial culture. In addition, the catch-all phrase private equity commonly covers a wide range of private investment activities including venture capital, growth capital, turnarounds and leveraged buyouts. (This study distinguishes between minority investments, as venture capital, and majority investments, as LBOs.) The structure of any given economy will often influence the relative activity level of early-stage versus mature company investment, and thus effect its overall riskreturn profile. People often talk of private equity as a young asset class, but US or British private equity are relative geriatrics compared with truly nascent markets elsewhere in the world. This matters because, in addition to fundamental considerations, the very fact of being a young private equity market will affect the performance data, as a greater number of young (closed ended) funds will still be in the process of creating value in the acquired assets. This has the effect of making younger markets look lower return, lower risk, when the opposite may actually be the case. Such are the pitfalls of blindly following the data, and what follows therefore strives to be objectively analytical and quantitative in its approach. But judgement must play a role. Where such subjectivity is unavoidable, we will always advertise this clearly. Geographical definition In this report, the method for geographical allocation of funds is based on geographical origin of the private equity fund investments. It should not be confused with a different method devised on the basis of location where the private equity firm is headquartered. This choice was made to allow for an explicit comparison of different geographical strategies in private equity investment decision making. Reporting currency For each geographical group of funds, the returns are calculated in terms of the local currency in which the cash distributions and residual valuations were reported 1. The reason for this approach is to abstract the exchange rate premium and risk from the performance measures analysed in this report. 6 efront Annual Report 2018

7 Measuring risk For the purposes of this study, we will define risk as the difference in performance between the best and worst performing funds in any given region. We will use two measures: one of extreme selection risk, which refers to the difference between the top and bottom 5% performers and most frequent selection risk, for the difference between the top and bottom quartiles. The former helps to assess overall selection risk, while the second neutralises any statistical outliers. Measuring return The closed ended nature of private equity funds also means return measurements should be multi-dimensional. For simplicity we will focus on the time sensitive internal rate of return, and the multiple of invested capital 2. Pooled average return indicators will be used to assess the performance of a given set of funds. Geographical definition In this report, the method for geographical allocation of funds is based on geographical origin of the private equity fund investments. It should not be confused with a different method devised on the basis of location where the private equity firm is headquartered. This choice was made to allow for an explicit comparison of different geographical strategies in private equity investment decision making. We wish to admit to our readers that geographical markets covered in this analysis vary in their size and recommend the focus on the comparison of markets that share similar size. 1 The currencies used are the following: British Pound for UK, Chinese Renminbi for China, Euro for Finland,, Germany,, Spain, Benelux, EEC & Russia, Nordic, Southern and Western Europe, i New Shekel for and US dollar for USA and Latin America. 2 Also known as total value to paid-in (TVPI) efront Annual Report

8 1. Main findings Private equity beyond the obvious Our initial analysis of private equity, as a whole, puts Benelux, Finland and China at the top of the risk-return spectrum. These are all markets with significant venture capital activity. But a further break down shows this to be a canard. If anything, Finnish venture capital drags down the risk-adjusted returns of private equity funds as they record a negative performance and present significant selection risks. Benelux VC funds are too recent to provide a meaningful risk-return profile. Meanwhile, it is not necessarily Chinese VC funds that are propping up its performance, but a lack of maturity of both VC and LBO funds in the region. There is a clear difference in maturity between European and emerging markets, but this does not always break down intuitively. Germany is a rather young private equity market despite being a mature industrialised nation. Its LBO and VC funds are still in the making. For other markets, relatively low IRRs tend to be a function of holding periods that is the case for the relatively low performance of French funds for instance. By contrast, Italian funds have a faster rotation of assets and show a particularly high performance in VC. has been through an accelerated path to development of its VC market, and its LBO funds record more conservative returns. As for Spain, it is in effect a lower risk-lower return market in VC and LBO. Clearly, jumping to conclusions is dangerous when analysing private equity, particularly on the basis of money returned or, worse still, IRRs. One must instead look at risks and returns, maturity of funds, time-to-liquidity and investment stages. Leverage buyouts The Nordics and Benelux are high risk-return LBO markets. The full value creation stories in younger LBO markets of China,, Germany and Southern European LBO markets have yet to show up in the data. There is also still scope for significant performance changes in, Spain and the US. 8 efront Annual Report 2018

9 Benelux The Benelux LBO market is at higher end of the risk-return spectrum. Both the pooled average IRR and the money multiple (1.85x) are top performers, and time-to-liquidity are short. Realised funds confirm this performance at 1.95x for vintage years , with fourfifths of assets realised. The MIRR of Benelux LBO funds stands at 5.48%, the highest of the sample. Benelux funds may appeal to investors able to bear significant selection risk. China At first glance, Chinese LBO funds offer attractive an IRR with an exceptionally low selection risk. In fact, the lower risk may be related to its lower maturity: only 53% of the returns have been realized so far for funds of vintage years 2008 to 2012 (collectively reporting a 1.73x). Thus, the high money multiple of the full sample (1.61x) is still largely in the making, due to long time-to-liquidity (+4 years). The jury is still out for China, notably to assess its overall risk and actual cash-on-cash performance. Eastern Europe and Russia The region s LBO market displays significant selection risk and lower performance (6.63% and 1.34x). Local LBO funds are still at a very early stage of their development: vintage years 2008 to 2012 have only realised 34% of assets. The average holding period 3.7 years is likely to increase and performance figures are also expected to change materially. Vintage years 2003 to 2007 reached a pooled average performance at 1.46x, currently realized at 63%. TVPI may therefore increase going forward, if not IRR. Nordics Nordic LBO markets are at the higher end of the risk-return spectrum. In terms of IRR, Finnish LBO funds rank behind Benelux and before the UK, with 16.73%. But by TVPI they come on top, with a pooled average performance of 1.96x. Finland has high extreme selection risk. Unsurprisingly, Nordic LBO funds rank high in terms of multiples (16.47% and 1.91x) and risk. Vintage years 2008 to 2012 generated a 13% and 1.58x return (realized at 70%). Interestingly, this was at par with Southern Europe (12.7% and 1.58x, realized at 63% only) and below Western Europe (17% and 1.82x, though realized at 58% only). Conditions might have become more favourable again to LBO investments in Nordic countries, which reached 2.02x with vintage years 2003 to French LBOs offer an attractive risk-return profile (10.24% and 1.51x), although more conservative than Western Europe in general (14.32% and 1.68x), and notably the UK (15.76% and 1.62x), and the US (12.17% and 1.56x). This is in line with LBO fund performance from vintage years 2003 to 2012, but French LBO funds take time to mature and might therefore see their overall performance change. Vintage years 2003 to 2007 (1.55x) have realised 86% of portfolios, with an average time-to-liquidity of 5.2 years. -British (1.61x) and Western European funds (1.62x) are 91% realised, and American funds (1.52x) are 81%. There is a clear relationship between maturity and performance and French funds have still to unlock some of their return potential. efront Annual Report

10 Germany and DACH It is difficult to draw conclusions from the performance of the DACH region LBO market, as it is less mature and has long time-toliquidity (more than four years). German LBO funds have, on paper, the worst risk-adjusted performance with an IRR of 5.87% and TVPI of 1.27x and the highest most frequent selection risk. However, funds mature slowly in the DACH region. Vintage years 2003 to 2007 (realized at 89%) generated a 1.42x TVPI and a 7.6% IRR as time-to-liquidity reached 4.8 years to 2012 vintage years are only 32% realised, the lowest level of the sample. This partially explains the 1.23x and 5% performance, notably as Germany has the second highest time-to-liquidity (4.2 years) for these vintage years. s LBO funds rank as the second worst in terms of performance in our sample (4.43% and 1.22x). They have the least attractive extreme selection riskreturn profile, although their most frequent selection risk-return profile is in line with the trend. This is puzzling for a country known for its entrepreneurial spirit and its achievements in the start-up world. However, most of the funds are still immature. LBO funds of vintage years 2008 to 2012, realized at 71%, record an IRR twice higher (9.9%) and a TVPI of 1.46x. Funds might still record additional exits and see their TVPI increase. As the market for LBO investments develop, the performance might be less erratic and increase. Active funds currently show a pooled average TVPI of 1.51x. Despite having short time-to-liquidity (3.2 years), Italian LBO funds do not have particularly high IRRs (3.73%). These conservative returns (1.19x) would normally command low levels of risk. Indeed, selecting Italian LBO funds is less risky than in Western Europe at large, but more than in Germany even though the latter has a higher TVPI (1.27x) and a lower maturity. Italian LBO funds have been through a steep learning curve. Vintage years were the only ones in the sample to lose money (0.9x and -1.8%) with a very long time-to-liquidity (5.9 years). This experience helped to reshape local LBO investing as funds of vintage years record a TVPI of 1.54x and an IRR of 14.4%, realized at 71%. Therefore, the aggregate performance of Italian LBO funds is weighted down by negative historical performance, obscuring better recent performance and return potential. Latin America Latin American LBO funds have the lowest est return, but also the lowest risk of the sample. Local funds use little to no financial leverage, which limits the risks and also the potential returns. The TVPI stands at 1.13x and the IRR at 1.97%. Given the limited number of active funds, the specific conditions of investment and the overall limitations of activity there, it is difficult to compare Latin American LBO funds with the rest of the peer group. Spain Despite long time-to-liquidity (5.5 years for vintage years ), Spanish LBO funds manage to generate relatively high IRRs (7.23%) with a fairly limited risk. Although Spanish funds are below the trend line in terms of riskadjusted returns, they provide investors with a 1.40x TVPI. This is in line with their historical 10 efront Annual Report 2018

11 performance of More recently ( ), Spanish funds reached a 1.61x (and 11.1% IRR), realized at 61% as local funds have the longest time-to-maturity (4.5 years). Assuming that active funds follow this path, the current aggregated performance figures might significantly understate the potential returns. Assuming that risks remain as limited as today, Spanish LBO funds might turn out to among the best if not the top risk-adjusted performers. UK variations. Over time, the US delivered a rather consistent performance: 1.52x in with vintage years (realized at 81%) and 1.54x with (realized at 62% only). Timeto-liquidity tends to be on the longer end of the spectrum, which means that IRRs are not necessarily the highest (8.15% for and 11.9% for ). Fully realized funds used to generate a 1.78x and a 15.07% IRR, but the US market is now much more developed and it is unlikely that performance will reach such levels going forward. UK LBO funds provide an attractive risk-return profile, above trend. On aggregate, the overall risk-return profile is more aggressive than, with a 15.76% IRR and a 1.62x TVPI. The TVPI is in line with past performance: 1.61x for vintage years (realized at 91%) and 1.7X for (realized at 71%). Holding periods used to be 5.4 years ( ) and decreased to 3.5 years ( ), but many funds are still holding assets in the latter case. Therefore, current IRR is expected to decrease, maybe not to the 9.21% of but closer to the current IRR of 11.91% of active British LBO funds. Although the UK is the most developed LBO market in Europe, it still delivers solid returns to investors able to select funds wisely. US At par with Western Europe when it comes to extreme selection risk, it shows a lower level of risk when it comes to most frequent selection risk. US LBO funds deliver more conservative return (TVPI of 1.56x) than Western Europe (1.68x). This might be related to the higher level of competition in the US, which are the most developed LBO market. Differences in vintage years in the sample might also explain certain efront Annual Report

12 Venture capital Countries with funds that invested capital during the late 1990 s boom and that avoided the crash of experienced the largest returns. VC funds in China generate exceptional performance combined with a low selection risk, but they are characterized by a low level of fund maturity. is at the higher end of the risk-return spectrum, while Spain is at the lowest end. Among fully realized funds, US VC funds dominates at the highest end of the risk-return spectrum. However, the UK offers the highest risk-return among active funds. Benelux Benelux VC funds exhibit a modest performance (0.05% and 1.0x) but with a moderate risk. This seems to be the consequence of the crash of on local VC fund performance. Indeed, active funds record a 1.17x and 3.41% performance, realized at 76.1%. As the market develops and mature, the relative positioning of Benelux VC funds will be clearer and more definitive conclusions could be drawn. China Chinese VC funds appear as outliers. A high level of IRR (11.53%) and TVPI (1.72x) combined with a low selection risk could appear as attractive. However, a low level of maturity (39% for active funds generating a performance of 12.69% and 1.75x) means that risk will increase and IRRs will decrease. It is therefore too early to draw conclusions. Eastern Europe and Russia Venture investing seems challenging as the performances of funds (0.19% and 1.01x) are associated with the second lowest most frequent selection risk in terms money multiple (after Benelux), largely due to the market s youth. To get a flavour of the potential performance of local VC funds, vintage years 2008 to 2012 provide a 1.17x and a 5.4% realized at 23%. It is early days for this market. Finland and Nordic countries Finland, the star of the LBO sector, is the laggard in venture capital. Performance of funds is negative in terms of IRR and TVPI (-3.49% and 0.84x) and risk profile appears as rather high, notably when it comes to the most frequent selection risk. The risk-return profile of Finnish VC funds could still evolve, as maturity is 59.2%. Finland appears as a regional outlier, as overall Nordic VC funds record relatively attractive performances (2.27% and 1.18x). Active funds stand currently at 6.11% and 1.53x, with a maturity of 53.4%. Fully realized funds generated a 14.86% and 1.87x performance, seemingly having fended off the effects of the period. Nordic funds seem to offer a compelling risk profile for investors with solid fund selection skills. The second largest European VC market is a reflection of the overall trend, with rather moderate returns (3.19% and 1.19x) and 12 efront Annual Report 2018

13 selection risks. Western Europe does better (5.32% and 1.34x), as well as the UK (4.95% and 1.37x). However, a look at active funds shows that the gap is narrowing. Fully realised French VC funds (3.45% and 1.15x) slightly outperformed their British peers (2.55% and 1.21x), but this dynamic seems to have reversed with active funds. seems to offer a relative stability of pooled average VC returns with a reasonable challenge in terms of fund selection. Germany and DACH German VC is a relative new-comer. The few realized funds have generated at 1.08x TVPI and a 1.55% IRR, weathering the crash. Active funds, realized at 32.9% stand at 1.01x and 0.12%. The overall low selection risk level is a testament to the lack of maturity of local VC funds. Contrasting with their Chinese and Finnish peers, it appears that German VC funds apply a rather conservative approach in the valuation of their portfolios and/or local start-ups did not experience the same ramp up in valuations. i VC experienced the period with significant losses, but more recent vintage years show an attractive potential. While VC funds of vintage years (realized at 35%) show a pooled average performance of 1.92x, vintage years of 2009 to 2015 declare a performance of 1.82x. Active funds record so far a 1.30x TVPI and a 4.30% IRR. Italian VC funds offer the highest risk-highest return profile of the sample in terms of IRR, adjusting for outliers. In terms of TVPI, the overall performance of Italian VC funds reaches 1.35x. Italian VC funds offer the best risk-adjusted performance when the most frequent selection risks are considered (ex: China). Fully realized funds generated a 1.09x TVPI and a 6.78% IRR. It could be tempting to assume that active funds are propelling overall returns on the back of a favourable ramp-up of start-up valuations. Indeed, active Italian VC funds record a 1.48x TVPI and 11.11% IRR. However, this performance has been 68.1% realised. Often overlooked in favour of more advanced European VC markets, might be attractive for investors willing to capture attractive performance with moderate levels of challenge in fund selection. Spain In terms of IRR, Spanish funds are at the lowest end of the risk-return spectrum. A 1.16x TVPI associated with a 2.67% IRR imply a rather long time-to-liquidity. In terms of TVPI, the risk-adjusted return of Spanish funds is conservative but relatively attractive, slightly above trend. Active Spanish VC funds record coherent returns at 1.21x and 3.56% and are realized at 55.3%. There is still room for improvement, albeit more in multiples than IRRs, given the 5.4 years average holding period. Spanish VC funds might be attractive for investors willing to invest in start-ups without investing too many resources in the selection. UK The largest European VC market is slightly above the risk-adjusted return trend line, with an IRR of 4.95% and a TVPI of 1.37x. Active funds record similar attractive performances with a 6.07% IRR and a TVPI of 1.42x (realized at 63.9%). The UK is, however, a challenging efront Annual Report

14 market, positioned at the higher end of the risk-return spectrum. The performance of fully realized funds, which therefore includes the crash, stands at 1.21x and 2.55%. Vintage years generated a comparatively high performance (8.8 and 1.71x) with an average holding period of 6.4 years and 78% maturity. Vintage years are just 49% realised, but already record a 1.58x TVPI with a current IRR of 14.13%. For investors equipped with the right skillset, British VC funds are a good entry point in Europe, providing higher absolute returns than but lower IRRs due to longer time-to-maturity. US A post-2001 assessment of US VC sets a more level playing field, and on this basis, they perform consistently well. Funds of vintage years 2003 to 2007 generated a multiple of 1.69x (realized at 70%) and an IRR of 9.3%. The following vintage years ( ) record a 1.73x TVPI (realized at 38%) and an IRR currently at 14.3%. The time-to-liquidity of US VC funds tends to be at the lower end of the spectrum: 5.9 years for vintage years 2003 to The US is at the top end of the risk-return profile, in line with the trend. Active funds are positioned as less challenging in terms of selection risk, but also less rewarding. However, the risk-adjusted return profile of US VC funds appears as more rewarding with current funds than historically, assuming that active funds realize their performance without significantly affecting selection risks. By far the largest VC market in the world, the US has continued to develop and in the process, seem to have improved the risk-reward for investors. 14 efront Annual Report 2018

15 2. Performance analysis of the major PE markets 2.1 All private equity funds Analysis using IRRs The major private equity markets broadly divided into two categories Group 1: IRRs above 10% This group of relative outperformers (11-14% IRRs) includes the developed nations of the US, UK, Finland and regions such as Western Europe, Benelux, and Nordic countries, plus China. The group is remarkably concentrated and coherent in terms of risk-return profile. From this group, Finland and China display extreme risk when using our extreme selection risk measure (the delta of the best and worst 5%). (Fig. 1). However, using the most frequent risk measure (the difference between the top and bottom quartiles), these two countries are in line with the rest of the group, though Finland is at the higher end of the risk spectrum. Benelux displays exceptionally high return and selection risk (Fig. 2). Fig. 1 Risk (5%) and return (IRR) analysis of private equity funds by geographical area Return vs. Risk (IRR) 16% 14% 12% China United Kingdom Nordic Benelux Western Europe United States Finland 10% R² = IRR 8% 6% 4% EEC & Russia Spain Southern Europe Germany 2% Latin America 0% 25% 35% 45% 55% 65% 75% 85% 95% Top 5% - Bottom 5% Spread Source: efront Insight, as of Q efront Annual Report

16 Group 2: IRRs 10% and less This lower return group (3-8%) also includes developed markets including, Germany,,, Spain, as well as Eastern Europe and Latin America. The group is more idiosyncratic when measured by extreme selection risk (Fig. 1), and the dispersion increases when looking at the most frequent selection risk (Fig. 2). These discrepancies could be partly a function of maturity differences as well as sample composition (vintage years and strategies). Fig. 2 - Risk (25%) and return (IRR) analysis of private equity funds by geographical area 16% Return vs. Risk (IRR) 14% 12% Western Europe Nordic United Kingdom United States Benelux China Finland 10% R² = IRR 8% 6% 4% Spain EEC & Russia Southern Europe Germany 2% Latin America 0% 5% 10% 15% 20% 25% 30% Top Quartile - Bottom Quartile Spread Source: efront Insight, as of Q Analysis using money multiples The two groups identified above are less visible on this ground IRRs are particularly time sensitive. To mitigate analytical biases associated with timing, we turn to TVPIs. The risk return relationship is clearest in Figure 4, which uses the money multiple return measure and the most frequent selection risk measure. The exceptions to this are China and the UK, which exhibit surprisingly high multiples (respectively 1.67x and 1.59x) for a rather limited (most frequent selection) risk spread. This might be due to a lower level of maturity, as fund spreads increase with their maturity. 16 efront Annual Report 2018

17 Germany and offer relatively low multiples (1.24x and 1.18x respectively) for a rather high most frequent selection risk. However, the composition of the sample and/or the maturity of the funds may play a role in the relative underperformance of these countries in the data. Once again, Finland is an outlier in terms of high risk and return. Overall, Western Europe offers an attractive risk-return profile. The US is slightly below the regression line, which suggests that selection risks are less well rewarded than in other geographies. Fig. 3 Risk (5%) and return (TVPI) analysis of private equity funds by geographical area 2.10 Return vs. Risk (TVPI) 1.90 Nordic 1.70 China Benelux Western Europe United Kingdom R² = United States Finland 1.50 TVPI 1.30 Spain EEC & Russia Southern Europe Germany Latin America Top 5% - Bottom 5% Spread Fig. 4 - Risk (25%) and return (TVPI) analysis of private equity funds by geographical area Return vs. Risk (TVPI) Nordic 1.70 China R² = Western Europe United Kingdom United States Benelux Finland TVPI 1.50 EEC & Russia Spain 1.30 Latin America Southern Europe Germany Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q efront Annual Report

18 2.2 Leverage buyout funds Analysis using IRRs Countries with more mature LBO markets yield higher returns Group 1: IRRs above 10%. The Benelux appears to offer the highest return but also the highest (most frequent selection) risk LBO market in the world (Fig. 5 and 6), with an IRR of 17.27%. Finland also belongs to this higher risk-higher return category and has the highest extreme selection risk. However, it seems to deliver a higher risk-adjusted IRR than the UK when the most frequent selection risk measure is used. Meanwhile, the UK ranks relatively favourably when the extreme selection risk is applied. Interestingly, the data suggests that private equity markets become more efficient in certain regions, the (most frequent selection) risk profile of their LBO market improves (Fig. 5 and 6). The US is a case in point. Although the country ranks almost at par with Western European LBOs in terms of extreme selection risk, it ranks lower in terms of most frequent selection risk. Fig. 5 Risk (5%) and return (IRR) analysis of LBO funds by geographical area LBO: Return vs. Risk (IRR) 20% 18% 16% United Kingdom Benelux Nordic Finland 14% Western Europe 12% United States R² = 0,2146 China IRR 10% 8% Spain 6% EEC & Russia Southern Europe Germany 4% 2% Latin America 0% 25% 35% 45% 55% 65% 75% 85% 95% Top 5% - Bottom 5% Spread Source: efront Insight, as of Q efront Annual Report 2018

19 Group 2: IRRs 10% and less The group of countries achieving single digit IRRs is composed of Spain, Germany,, and Eastern Europe. The relatively high spread associated with lower returns might look less attractive. Fig. 6 - Risk (25%) and return (IRR) analysis of LBO funds by geographical area LBO: Return vs. Risk (IRR) 20% 18% 16% United Kingdom Benelux Nordic Finland 14% Western Europe 12% United States R² = 0,2146 China IRR 10% 8% Spain 6% EEC & Russia Southern Europe Germany 4% 2% Latin America 0% 25% 35% 45% 55% 65% 75% 85% 95% Top 5% - Bottom 5% Spread Source: efront Insight, as of Q However, some caution must be exercised in extrapolating this to manager quality: LBOs in these regions may suffer from less favourable refinancing environments (dividend recapitalisations are less frequent) as well as longer time-to-liquidity (see Fig. 7). Moreover, the underlying vintage years, as well as the maturity of funds might play a role (Figs. 7 and 8). The importance of timing The IRR is particularly duplicitous when assessing LBOs since they are sensitive both to the timing of cash flows and to the overall duration of investments (measured by the average time-to-liquidity in Fig. 7). In addition, LBO funds can distribute early thanks to dividend recapitalisations, in effect creating distortions in the analysis of performance. efront Annual Report

20 Fig. 7 Time-to-liquidity of LBO funds by geographical area LBO: Time to Liquidity 5,0 4,5 Time to Liquidity (years) 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0 China Spain United Kingdom United States Benelux DACH EEC & Russia Nordic Southern Europe Western Europe Geographical Area Source: efront Insight, as of Q Historically, Spanish LBOs have had the longest time-to-liquidity (4.5 years), as well as China,, and the DACH region, which were above 4 years (Fig. 7). The shortest holding periods have been in (3.2 years), (3.4 years) and the UK (3.5 years). The current holding period for active vintage years ( ) show that and the Benelux LBO funds are indeed slower at exiting investments. Spain, China, the US and Eastern Europe might look as if they have shorter time-to-liquidity. However, as the funds are still investing, it might also be that they are in the process of catching up with their historical average. The duration of holding periods also has an influence on the analysis of performances of active funds. Fig. 8 provides a snapshot of relatively mature funds, which have finished investing and seeking to exit their remaining investments. While Benelux has distributed 81% of its total capital paid-in, Germany, Austria, and Switzerland have only distributed 32%. This difference of maturity explains why the former ranks very high in terms of performance and risk; and the second is at the lower end. Not only are IRRs sensitive to the ability of Benelux funds to have a high rotation of assets with limited duration, but the money multiple records the full value created by these managers. By contrast, DACH funds have yet to unlock the full value created in their portfolio companies. 20 efront Annual Report 2018

21 Fig. 8 Return and maturity of LBO funds by geographical area (vintage years ) LBO: TVPI vs. Maturity ( ) 2 1,9 Benelux 1,8 Western Europe United Kingdom 1,7 China Spain TVPI 1,6 1,5 R² = 0,498 Southern Europe United States Nordic 1,4 1,3 1,2 DACH EEC & Russia 1,1 1 20% 30% 40% 50% 60% 70% 80% 90% Maturity Source: efront Insight, as of Q The performance of British, i, Italian and Nordic LBO funds includes around 70% or more of realised value. This means that their overall performance is unlikely to shift dramatically for these vintage years. There is still room for significant change in, Spain, and the US. As for China and Western Europe, the performance remains still significantly in the making. Analysis using money multiples Strong correlation between risk and TVPIs There is a strong correlation between risk and money multiple return (Fig. 9 and 10) in contrast to IRRs which have been skewed by dividend recaps, which are influenced by the timing of cash flows. Adjusting for this fails to explain away the position of Finland and Benelux on the risk-return spectrum. The US, the UK,, and Western Europe offer a more conservative but attractive, most frequent selection risk-return profile. Spain, Germany, Eastern Europe, and Latin America look less attractive. China and appear as statistical outliers with particularly low most frequent selection risks. In case of China, a lower maturity might explain its position. efront Annual Report

22 Fig. 9 Risk (5%) and return (TVPI) analysis of LBO funds by geographical area LBO: Return v s. Risk (TVPI) Benelux Nordic Finland 1.70 United Kingdom China Western Europe R² = United States TVPI Spain Southern Europe Latin America EEC & Russia Germany Top 5% - Bottom 5% Spread Fig Risk (25%) and return (TVPI) analysis of LBO funds by geographical area LBO: Return v s. Risk (TVPI) Nordic Benelux R² = Finland 1.70 China Western Europe United States United Kingdom 1.50 TVPI 1.30 EEC & Russia Spain Southern Europe 1.10 Latin America Germany Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q efront Annual Report 2018

23 Performance dispersion greater for active LBO funds The relationship between money multiple and most frequent selection risk is strong, as shown by Fig 11 s analysis of fully realized funds. Nordic LBO funds deliver strong performance but investing in the region also carries significant selection risks. Manager selection is also a major factor in the US, which sits below the trend line. Meanwhile Western Europe sits on the trend line, thus delivering the overall return expected for the level of risk taken. The UK and are above the trend line, therefore delivering higher returns for the level of risk taken by investors. provides a more conservative risk-return profile, while the UK is more aggressive. Southern Europe delivered a negative pooled average return. Its fund managers were still in the learning curve and therefore did not differentiate strongly from each other at that time. Fig Risk (25%) and return (TVPI) analysis of fully realized LBO funds by geographical area 3.50 LBO Inactive: Return vs. Risk (TVPI) 3.00 Nordic 2.50 R² = TVPI 1.50 United Kingdom Western Europe United States 1.00 Southern Europe Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q The picture is less clear with active funds (Fig. 12), due to differences in terms of maturity and time-to-liquidity., China, Spain, Germany, Eastern Europe, as well as for Western Europe to a certain extent, are still in the making. It is therefore difficult to draw any conclusion on their absolute and relative risk-adjusted performance. efront Annual Report

24 , the US, and the UK seem to be on the trend line, which is in line with their historical performance. However, the absolute level of performance of their active funds is below the historical performance of their realized funds. This phenomenon affects the three most developed private equity markets in a similar way, and could be the result of two factors. Either significant reserves of value in portfolios have still to be unlocked and/or this reflects a trend towards lower LBO performance in these countries. Nordic and Benelux LBO funds seem to confirm their position at the higher end of the riskreturn spectrum. 2,00 Fig Risk (25%) and return (TVPI) analysis of fully realized LBO funds by geographical area LBO Active: Return vs. Risk (TVPI) 1,90 1,80 Finland Nordic Benelux 1,70 Western Europe TVPI 1,60 1,50 1,40 China United States Spain United Kingdom 1,30 EEC & Russia Southern Europe Germany 1,20 1,10 1,00 0,40 0,50 0,60 0,70 0,80 0,90 1,00 Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q efront Annual Report 2018

25 2.3 Venture Capital Funds Analysis using IRRs The largest returns in the countries with funds that invested capital during the late 90 s boom and that avoided the crash of Italian VC funds rank at the top end of the risk-return spectrum (Fig. 13 and 14), with an IRR of 10.11%. This applies to both most frequent and extreme selection risk. This could come as a surprise, as usually more advanced markets such as the US, the UK, and are mentioned as the most attractive destinations for VC investing. Indeed, the US tops the ranking in terms of return, but its extreme and most frequent selection risk levels are lower than in. If US VC IRRs look particularly attractive, it is probably because they capture the legacy of the boom years of China also ranks particularly high in terms of IRRs. However, it has the third lowest extreme selection risk and it is a notch above the US as for the most frequent selection risk. This unusual risk profile is a consequence of the lack of maturity of Chinese funds (39.1%). Indeed, there is a direct relationship between maturity and risk (Fig. 15), and therefore Chinese VC funds appear as statistical outliers. 16% Fig. 13 Risk (5%) and return (IRR) analysis of VC funds by geographical area VC: Return vs. Risk (IRR) 14% United States 12% China 10% 8% R² = IRR 6% Western Europe United Kingdom 4% Southern Europe Spain 2% Nordic EEC & Russia Germany 0% 20% 30% Benelux 40% 50% 60% 70% 80% 90% 100% 110% -2% -4% Finland -6% Top 5% - Bottom 5% Spread Sources: efront Insight, as of Q With the exception of, the US and China, IRRs look particularly low: none of the other countries plotted exceeds 5% (Fig. 13). This is far below the usual preferred rate of return ( hurdle rate ) of 8% routinely requested by fund investors. efront Annual Report

26 Fig Risk (25%) and return (IRR) analysis of VC funds by geographical area 16% VC: Return vs. Risk (IRR) 14% United States 12% China 10% 8% IRR R² = % Western Europe Southern Europe United Kingdom 4% Spain 2% Nordic Germany Benelux EEC & Russia 0% 5% 7% 9% 11% 13% 15% 17% 19% 21% 23% 25% -2% -4% Finland -6% Top Quartile - Bottom Quartile Spread However, IRRs in venture capital must be interpreted with caution. First, most of the countries are European and never experienced the boom of the late 1990s to compensate for the crash of China was spared this crash as its venture capital industry emerged after. Fig. 15 Risk (25%) and maturity analysis of active VC funds by geographical area 0.95 VC Active: Maturity vs. Risk (TVPI) United Kingdom 0.85 Top Quartile - Bottom Quartile Spread EEC & Russia Germany Western Europe China & Hong Kong Nordic R² = Spain Finland United States Southern Europe Benelux Maturity Sources: efront Insight, as of Q efront Annual Report 2018

27 This also explains the relative concentration of extreme selection risk (Fig. 13): as most European VC funds invested at the top of the cycle, none of them could generate high returns. The only exception is Germany. As a consequence, the dispersion of returns applies more for top and bottom quartile fund managers (Fig. 14). Even though the pooled average performance of European VC funds (except ) does not look attractive, top quartile fund managers have in effect generated performance. This includes Finland, despite a pooled average of -3.49%. Second, start-up investments require a much longer holding period to mature. In that respect, venture capital is largely a seed and early-stage financing game, funding startups at inception and supporting them during the following five to nine years (Fig. 16). For example, funds of vintage years 2003 to 2007 have a time-to-liquidity of at least 5.5 years, the UK exceeding six years and China seven. This time-to-liquidity is expected to increase, as maturity levels do not exceed 80%, with the UK at 78% and China at 61%. Therefore, venture capital investing is more a money multiple than an IRR play. Even then, multiples are sensitive to portfolio appraisals by fund managers, as rounds of financing can artificially increase the value of portfolio companies, while the actual exit might happen at a significantly lower valuation. Fig. 16 Time-to-liquidity of VC funds by geographical area VC: Time to Liquidity Time to Liquidity China United Kingdom United States Southern Europe Western Europe Country Sources: efront Insight, as of Q efront Annual Report

28 Analysis using money multiples China and preserve their above the trend performance Excluding Chinese VC funds, the relationship between risk and return is strong and significant (with an r-square of 0.61 for extreme risk selection and 0.4 for most frequent selection risk). The US rank at the higher end of the risk-return spectrum, due to performance of the late 1990s (Fig. 17 and 18). confirms its ranking with at 1.35x and a risk-return profile above trend but appears as far less risky than when IRRs are used. British VC funds offer an attractive but challenging risk-return profile: they are on trend, but at the higher end of the risk spectrum (Fig. 18). Benelux VC funds are too young to provide meaningful results, as are Eastern Europe and Russia. This might also be the case for Germany and perhaps Finland. At first glance, i VC funds provide conservative returns (1.17x) for rather high extreme selection and most frequent selection risks. This might be due to the combination of a rather recent emergence of local VC funds and an accelerated development. Vintage years 2003 to 2007 (Fig. 16) generated a 1.21x (realized at 66%). Vintage years 2008 to 2012 have so far a 1.92x performance (realized at 35%). Although other countries have also recorded a progression of their pooled average performance during the same time frames, none has at such scale. has seen an increase from 1.27x to 1.44x, the US from 1.69x to 1.73x. China and the UK recorded a decrease. Overall, Western Europe has seen a progression from 1.49x to 1.65x. Fig. 17 Risk (5%) and return (TVPI) analysis of VC funds by geographical area VC: Return v s. Risk (TVPI) 1.80 China 1.60 R² = United States 1.40 United Kingdom Western Europe Southern Europe TVPI 1.20 Nordic Spain 1.00 EEC & Russia Benelux Germany 0.80 Finland Top 5% - Bottom 5% Spread Sources: efront Insight, as of Q efront Annual Report 2018

29 Fig Risk (25%) and return (TVPI) analysis of VC funds by geographical area 1.80 VC: Return v s. Risk (TVPI) China 1.60 United States R² = Western Europe United Kingdom TVPI 1.20 Southern Europe Spain Nordic 1.00 Benelux EEC & Russia Germany 0.80 Finland Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q Active funds experience acceleration in performance Fully realized funds (Fig. 19) confirm this perspective: i funds had a rough start, and collectively lost capital (0.59x), while active funds (Fig. 20) are on track to generate an attractive performance (1.30x), on-trend when most frequent selection risk is considered. Finnish VC funds, however, seem to suffer from difficult conditions. Fully realized funds generated a 0.85x performance, and active funds stand at 0.83x. This might hint at persistently high local valuations. seems to remain stable and gravitate around the trend line. The jury is still out for German VC funds, which are largely unrealized. Not only Italian funds have seen an increase of performance from 1.09x to 1.48x but the latter is 68.1% realised. Italian VC funds offer an attractive risk-return profile. Surprisingly, as the UK is the most developed European venture capital market, the risk-return profile British VC funds remains below or exactly on the trend line (Fig. 19 and 20), despite an increase of performance from 1.21x to 1.42x (realized at 63.9%). Therefore, if British VC funds provide attractive returns, selecting the right funds is quite challenging. Not surprisingly, the US has recorded a decrease in performance from 1.82x to 1.49x, as active funds no longer include the exceptional vintage years of the 1990s. However, US VC funds still provide attractive risk-adjusted returns, only outranked by Italian and Nordic VC funds (Fig. 20). The development of the American market has shifted local VC funds from the highest risk-return profile to a more moderate one, but also more attractive when most frequent risk selection is considered. One may speculate whether the UK, which has replaced the US at the highest risk-highest return position will follow the same path. efront Annual Report

30 Fig Risk (25%) and return (TVPI) analysis of fully realized VC funds by geographical area 2.00 VC Inactiv e: Return v s. Risk (TVPI) 1.80 United States 1.60 R² = TVPI 1.20 Germany United Kingdom 1.00 Southern Europe Western Europe 0.80 Nordic Finland Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q Fig Risk (25%) and return (TVPI) analysis of active VC funds by geographical area 2.00 VC Activ e: Return v s. Risk (TVPI) 1.80 China 1.60 Nordic United States 1.40 Western Europe United Kingdom TVPI Southern Europe R² = Spain Benelux 1.00 EEC & Russia Germany 0.80 Finland Top Quartile - Bottom Quartile Spread Sources: efront Insight, as of Q efront Annual Report 2018

31 Methodology Fig. 1 is based on Internal rate of return (IRR) as a measure of return performance and the difference between the average IRR of the top 5% performers and the average IRR of the bottom 5% performers as a measure of fund selection risk. For each geographical group of funds of all the vintages available, the average IRR is calculated. This pool of funds includes both active and liquidated funds. In terms of the investment strategy, both LBO and VC funds are represented in the figure. The purpose is to exhibit the risk-return profile of private equity investments in each geographical region. Fig. 2 is based on Internal rate of return (IRR) as a measure of return performance and the difference between the average IRR of the top quartile performer and the average IRR of the bottom quartile performer as a measure of fund selection risk. For each geographical group of funds of all the vintages available, the average IRR is calculated. This pool of funds includes both active and liquidated funds. In terms of the investment strategy, both LBO and VC funds are represented in the figure. The purpose is to exhibit the riskreturn profile of private equity investments in each geographical region. Fig. 3 is based on a multiple of invested capital (total value to paid-in, TVPI), defined as the sum of capital distributed and the residual net asset values, in relation to the total amount of capital paid in to the fund up to date. TVPI is used as a measure of the performance of the funds, whereas the difference between the average TVPI of the top 5% performers and the average TVPI of the bottom 5% performers is used as a measure of fund selection risk. For each geographical group of funds of all the vintages available, the average TVPI is calculated. This pool of funds includes both active and liquidated funds. In terms of the investment strategy, both LBO and VC funds are represented in the figure. The purpose is to exhibit the risk-return profile of private equity investments in each geographical region. Fig. 4 is based on a multiple of invested capital (total value to paid-in, TVPI), defined as the sum of capital distributed and the residual net asset values, in relation to the total amount of capital paid in to the fund up to date. TVPI is used as a measure of the performance of the funds, whereas the difference between the TVPI of the top quartile performer and the TVPI of the bottom quartile performer is used as a measure of fund selection risk. For each geographical group of funds of all the vintages available, the average TVPI is calculated. This pool of funds includes both active and liquidated funds. In terms of the investment strategy, both LBO and VC funds are represented in the figure. The purpose is to exhibit the riskreturn profile of private equity investments in each geographical region. Fig. 5 is based on Internal rate of return (IRR) as a measure of return performance and the difference between the average IRR of the top 5% performers and the average IRR of the bottom 5% performers as a measure of fund selection risk. For each geographical group of funds of all the vintages available, the average IRR is calculated. This pool of funds includes both active and liquidated efront Annual Report