HarbourVest Global Private Equity — Secured liquidity

In the 10 months between the last reporting period (H120, ending 31 July 2019) and end-May 2020, HVPE’s NAV declined 0.7% in total return terms and ended the period with an NAV of US$25.62 per share. Fees and operating expenses were higher year-on-year and lowered the NAV per share by US$0.31 versus US$0.23 in the corresponding period of the previous year, which we understand was due to the lower interest income on the lower cash position as well as higher performance fees at the HVPE level.

HVPE’s performance was broadly in line with its benchmark (FTSE All-World posted an 0.8% decrease) and its PE-focused peers (LPX Direct NAV posted an 1.5% decrease). We need to note that HVPE’s NAV is estimated monthly, based on available underlying valuations, which is adjusted periodically on the release of interim and annual financial reports once all data from underlying funds are collected. We note that the periodicity of estimating the value of underlying investments and a resulting valuation lag is a characteristic shared by most entities in the PE space.

However, the end-May 2020 NAV is already based mostly (91% of the portfolio value) on end-March 2020 valuations and therefore to a significant extent captures the coronavirus-induced market downturn of March, resulting in a downward revaluation of US$1.75 per share versus the April NAV. Based on our discussions with management, the collected valuations at end-March reduced the NAV by more than US$2.00 per share, which was partially offset by the subsequent broader market rebound seen in publicly traded assets within the portfolio.

Solid historical NAV performance throughout the cycle

HVPE has a long track record of sustainable NAV returns proving the expertise of its investment manager, HarbourVest Partners, in selecting high-quality underlying general partners (GP) and co-investment opportunities. Since inception in December 2007 (shortly before the onset of the GFC) it has generated a 7.4% NAV TR per year to May 2020 (MSCI AW TR 4.4% per year) and achieved a blended realised exit multiple of 2.13x (buyout 2.31x; venture capital 2.01x) across 6,751 fully realised investments in the period since inception to end-July 2019 (date of last available data).

HVPE withstood the GFC relatively well

It is also worth noting that during the GFC, HVPE outperformed the broad public market. During the fiscal year ending January 2009 (shortly after the market-wide sell-off) HVPE’s NAV declined 27% y-o-y, which compares to a 42% decline for the MSCI World Index. It took HVPE two fiscal years to fully recover in NAV terms. Furthermore, its four-year NAV performance over the crisis (January 2008 to January 2012) was 9.9%, compared to a 4.3% decrease for MSCI World. It is important to note that due to its exposure to private investments, HVPE enjoys a more resilient NAV performance, similar to its peers. In addition to the above-mentioned valuation lag, we believe that companies backed by solid PE funds withstand downturns relatively better as they tend to have better access to capital (and in some cases value-creation expertise). Second, PE funds have wider buying opportunities to pursue in a distressed environment. In this context, we note that (as highlighted by McKinsey in its recent study), the EV/EBITDA (two-year trailing average EBITDA) multiple decrease was less pronounced in US buyouts, with transaction multiples falling from 9.4x in 2007 to 8.8x in 2010, than in public assets (Russell 2000 index) declining from 11.6x to 9.8x, respectively.

Discount narrowing due to share price rebound and NAV update

In recent months, the lagged portfolio valuation coupled with HVPE’s significant share price decline (triggered by the broader market sell-off) led to the NAV discount widening to a record high of 60% on 19 March 2020. Following the share price rebound on the back of the broader market recovery as well as the NAV update with end-March valuations, the discount now stands at 25% compared to a five-year average of 21%. We note that the discount comparability during the GFC is complicated by the fact that HVPE initially traded with a double-digit premium to its NAV due to the very low stock liquidity on Euronext Amsterdam. Nevertheless, after normalisation of trading, the stock continued to trade at a wide discount of 38% on average in the years 2010 to 2012.

Compared to the GFC, which HVPE withstood relatively well (see above), the current portfolio is larger and more diversified as it consistently grew over the last 12 years from an NAV of US$0.8bn in September 2008 to US$2.0bn at end-May 2020. In our view HVPE’s liquidity position is at a safe level, despite the fact that coverage ratios are lower than at the brink of the previous crisis.

HVPE’s coverage ratio tends to be below direct PE peers due to longer investment cycles

HVPE’s total investment pipeline at end-May 2020 was US$1.7bn (slightly up since July 2019). This compares to US$214.7m of liquid resources held at the HVPE level, which is further supported by the US$400m remaining undrawn credit line. The above translates into a commitment coverage ratio of 36%. Due to its fund-of-funds structure, HVPE’s investment cycle is longer than that of many quoted PE peers and thus requires a relatively high level of unfunded commitments to remain fully invested. Its coverage ratio may therefore appear relatively low while not necessarily being associated with greater risk. Still, we note the current coverage ratio is below its historical levels of 49% on average over the last three years, or 67% in September 2008.

The above commitments include capital committed to HarbourVest funds that have not yet allocated capital to GPs. After excluding these, the allocated commitments amounted to US$1.4bn versus US$1.2bn at end-July 2019. The corresponding allocated coverage ratio is 44% (vs three-year average at 67% and 109% in September 2008). The comparatively lower coverage ratios versus historical levels are partially explained by the increasing diversification of HVPE’s portfolio and the need for a hefty investment pipeline to be fully invested. To provide better clarity for investors, HVPE also discloses the rolling coverage ratio, which only takes into account commitments expected to be drawn over the next three years. The ratio stood at 63% at end-May 2020 and has fallen 13pp since March 2020 due to higher drawdowns, as described below. Over 2019 the rolling coverage ratio oscillated around 76%.

Liquidity supported by US$600m open credit line of which US$400m is still undrawn

HVPE has a US$600m credit line provided by Mitsubishi UFG and Credit Suisse, which was extended from US$500m in December 2018. In May HVPE drew an initial tranche of US$200m to strengthen its short-term liquidity, as management expects a further mismatch between drawdowns and distributions in the coming months. It is worth noting that HarbourVest funds use leverage to delay and aggregate capital calls from limited partners, including HVPE. This embedded leverage attributable to HVPE was US$354.6m at end-May 2020 (17.3% of the NAV), up 10% since end-H120 in July 2019 when it was US$323.1m (16.2% of the NAV).

The company highlights that with the remaining US$400m undrawn credit facility, it is well placed to continue investing as planned. Although the pace of drawdowns and the time the M&A market will be required to return to normal activity is difficult to forecast, we present some additional approaches to assessing HVPE’s liquidity. Assuming the commitments that are expected to be drawn over three years are drawn equally in each year, HVPE’s resources cover investments over the next 12 months almost twofold without any realisations. If we assume the first-year drawdown is equal to the current embedded leverage at HarbourVest funds, it would imply that HVPE’s resources cover 179% of the assumed drawdown. The calculations do not include ongoing charges, but we note that expenses amounted to c US$11m per year (c 0.8% of NAV) on average over the last five years, while cash distributions averaged US$330m per year during the period. That said, the initial drawdown of US$200m will increase the ongoing costs by c US$3m according to our calculations, as the interest costs will replace the non-utilisation fee on the drawn amount. The credit facility limits HVPE’s indebtedness to 35% of assets, which according to our estimates allows for full drawdown even assuming an NAV deterioration in line with the most severe period of the GFC (down 27%).

Realisations continue even during market turmoil

Interestingly, during the GFC, while HVPE was a net investor in the fiscal years ended January 2009 and 2010, realisations did not fully stop. The net investment (drawdowns compared to distributions) amounted to 5.2% and 2.7% of the NAV at the beginning of respective periods. This compares to net realisations of 5.6% pa on average over the last five years and the size of HVPE’s credit facility to NAV at c 29%. HVPE does not pay dividends and it has not returned any capital to shareholders since FY15. Consequently, we believe that despite coverage ratios that may seem low at first glance, HVPE’s liquidity position is indeed secure.

Over the period since July 2019, HVPE has been a net investor with US$326m called by HarbourVest funds (adjusted for the US$42m commitments returned from the real assets vehicle), while realisations amounted to US$220m. This was largely driven by activity in the most recent months, as distributions to HVPE have come to a standstill during the coronavirus crisis, while PE funds have increasingly called for capital. We understand this may be due, in part, to underlying GPs calling for capital to support their existing portfolio companies and take advantage of the opportunities to deploy capital at lower valuations. In the year to date, this resulted in net outflows at the HVPE level of US$147m (or 6.8% of end-December 2019 NAV). While there were no distributions to HVPE during April and May 2020, the realisations within underlying portfolios did not fully stop. In April and May, 30 M&A and five IPO transactions were performed (resulting in proceeds from GPs to HarbourVest funds), of which the largest was the disposal of Zobele (agreed in February 2020) with US$2.7m distribution attributable to HVPE. Nevertheless, the 17 liquidity events of May 2020 compare to 29 per month on average throughout 2019. Until the pause in commitments announced on 21 May 2020, the company committed capital to new HarbourVest funds at a steady pace, with US$300m committed over the last 10 months (vs US$460m in the 10 months ending May 2019), including US$50m committed to a global secondary fund in March 2020.

The recent investment activity was largely in line with HVPE’s long-term strategy and targeted portfolio exposures, with 72% deployed into buyout funds (with its exposure at end-May 2020 still slightly below the 60% target) and about three-quarters in US-based funds (vs 58% actual and 65% targeted). Due to its multi-layer structure HVPE does not have the ability to quickly implement large changes to its portfolio structure and increase exposure to arising megatrends. In contrast, this also protects the trust against excessive exposure to ‘hot’ and overrated investment themes. That said, the portfolio remains broadly similar since its inception in 2007 in terms of sector and geographic exposure.

HVPE’s portfolio is well diversified among GPs, vintages, sectors and underlying companies. HVPE is invested in over 50 HarbourVest funds that have access to over 8,500 GPs globally. At end-July 2019 the top 10 largest GPs within HVPE’s portfolio managed 22.9% of its capital. While it may happen that amid current market turmoil some GPs will need to sell investments to meet their liquidity requirements, the broad diversification of HVPE’s portfolio works as a limiting factor in our view. We note that ultimately HVPE had exposure to 9,160 underlying companies at end-July 2019, compared to 5,705 in January 2009.

In terms of split by industry, we note that HVPE’s exposure to sectors that are likely to be the most and the least resilient in the current market environment is comparable with the broader market (as illustrated by MSCI World). It has a combined exposure of 46% to sectors that are more resilient to the current economic downturn, including tech and software, medical and biotech, and media and telecoms (vs 44% for IT, healthcare and communication services in case of MSCI World), while its exposure to industrial and transport and consumer is 26% (vs 30% for MSCI World covering consumer discretionary and staples, as well as industrials). We also note that its exposure to tech, software, media and biotech/healthcare sector is slightly below the levels on the start of the GFC in 2008 (57% at September 2008).

The infrastructure assets exposure constitutes an additional differentiating factor from HVPE’s UK listed PE fund-of-fund peers, which either have no explicitly stated direct exposure to real assets or it is limited to a low single-digit share of their portfolios (although we acknowledge their actual exposure on a look-through basis may be higher). They also may constitute a good inflation hedge in the case of inflation-linked or correlated revenues. HVPE’s investment manager highlighted in the past that it has identified good opportunities in this space associated with a capital maturity cliff, which forces some owners of good-quality assets to sell or restructure them. We note that HVPE’s mezzanine/real assets made up 15% of the portfolio at end-July 2019 (11% at end-May 2020) and included a single GP exposure to Corsair Capital, representing 6.7% of its portfolio, in which HVPE invested in March 2019. This included three assets:

In our peer comparison table (Exhibit 8) we include six other members of the AIC Private Equity sector that also have portfolios mostly managed by third-party GPs. NAV total returns are in sterling terms and HVPE’s performance is ahead of the peer group averages over one, three, five and 10 years. HVPE trades at a discount in line with the peer average and its ongoing charge is comparable with the average within the group. The peer companies in the PE sector are subject to a lag in NAV performance just like HVPE (as described earlier) to various extents, as well as reporting periodicity of their respective NAVs. At the same time, as discussed above, HVPE’s May NAV already reflects most of the Q1 impact. Most of the peer companies incorporated the impact of market downturn from March on the majority of their portfolios as well, despite the fact that only HVPE and Standard Life Private Equity reported the NAV as at end-May. The exception is BMO Private Equity, where only 19% of portfolio reflects the decline (with average decrease of 7% in March). While the portfolio valuation of Pantheon International mostly reflects end-2019 data, the manager recently made provisions for expected losses at 7.8% of NAV, which is reflected in its performance. The remaining companies incorporated end-March valuations in the majority (77–98%) of underlying valuations. We believe the market possibly expects a degree of further deterioration in NAVs. This is clearly reflected in the high discounts to NAV, as the average of the peer group is now 29% compared to 17% when we published our previous report in September 2019. HVPE does not pay dividends, compared to the trailing yield of 3% by peers on average, but we note the dividend pay-out capacity is under pressure across the sector amid liquidity shortages throughout portfolio companies.