Cannes, France) - Measuringriskin private markets · 2020. 2. 4. · ¡ Pro-cyclicality ¡...
Transcript of Cannes, France) - Measuringriskin private markets · 2020. 2. 4. · ¡ Pro-cyclicality ¡...
Measuring risk in private markets
Prof Dr Cyril Demaria
IPEM // Cannes // January 30, 2020
2
20 years of experience in private markets
• Research: Executive director in charge of private markets research, UBS WM CIO
• Investments: venture capital fund manager, former CIO of Tiaré Investment Management AG
• Advisor to the President of CASDEN-Banque Populaire, a Swiss pension fund and a MFO
• Expert for SECA, France Invest, InvestEurope and CAIA
Six books on private equity, among which:
• Introduction to private equity, Wiley, 2020, 3d ed. (trans.: Portuguese, Spanish and Mandarin)
• Asset allocation and private markets, Wiley, forthcoming 2020
• Private equity fund investments, Palgrave, 2015
Prof Dr Cyril DemariaPartner, Head of Private Markets, Wellershoff & Partners
3
AgendaMeasuring risk in private markets
1. Asset price volatility is irrelevant to measure risk
2. What matters: potential losses
3. Strategy- and execution-related risk
4. Selection risk
5. What about... liquidity risk?
6. Conclusion
Asset price volatility is irrelevant to measure risk
Not only in private markets, but also in general
5
Used as a measure of risk, asset price volatility assumes:¡ Frequent and seamless tradability (so-called « liquidity »)¡ Normal distributions of returns (to run simulations such as Monte-Carlo)è Not applicable to private marketsè Instruments used for listed assets are not relevant (nor Sharpe ratios, etc.)
Private market assets are « less liquid » by construction¡ The liquidity horizon is the third dimension of investing (next to risk & return)¡ Liquidity horizons are predictable: the variability around them is the actual risk
è Focus on what matters to investors: losses of capital and variability of results¡ How much capital could I loose? (simplified value-at-risk assessment)¡ How do returns evolve due to macro and other factors? (strategy assessment)¡ What is my selection risk? (skillset assessment)¡ What is the actual evolution versus the plan? (execution & liquidity risk)¡ (Other, yet to be defined: credit risk for leveraged strategies?)
Volatility of asset prices is irrelevant to measure risk
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
6
PE: less volatile than the stock exchange
• Since 2007, LBO and Growth capital funds apply mark-to-market rules• VC, distressed debt and turn-around capital keep the historical cost ruleè Volatility is not a relevant measure of risk in PE (illiquidity + fund shield)
Source: Preqin, S&P, Wellershoff & Partners, 2018.
1
324 285 271 349 419 496630
837 743904
1106 12001382
1512 1552 1635 1744
2040
-300
200
700
1200
1700
2200
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
USDbn
Valueofcompaniesinportfolio(unrealizedvalue) S&P500
Compared quarterly evolution of major stock indexes and PE NAV
S&P 500 2008 = 100
Historicalcost FMV
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
7
Performance distribution is not normal
Sample: 5000 direct US investments. 300 European VC funds and 618 funds of funds.Source: Weidig & Mathonet/Thomson VentureXpert, 2003, Wellershoff & Partners/Cambridge
Associates, 2017.
Dispersion of returns in VC direct (US), funds (Europe) and funds of funds (generalist)
0%
10%
20%
30%
40%
50%
60%
70% 0
0.2
5
0.5
0.7
5
1
1.2
5
1.5
1.7
5
2
2.2
5
2.5
2.7
5
3
3.2
5
3.5
3.7
5
4
4.2
5
4.5
4.7
5
5
Pro
bab
ility
Distributed to Paid-In Directs Funds Funds of funds
Fund-of-funds are highly concentrated also skewed.
Direct highly skewed, 30% of investments are total losses!
Extreme profits for direct.Healthy profits for fund.
Funds are less skewed and widespread.
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
8Source : Cambridge Associates, Wellershoff & Partners, 2017.
The third dimension of investment: liquidity horizonDispersion of PE strategies based on their performance and duration
• PE: time-to-liquidity is generally of 3.5 to 5.5 years• Possibility to assess the historical and current variation around the reference pointè Portfolio construction can benefit from using this dimension
Wellershoff & Partners– C. Demaria – 29/1/2020Material under copyright – Do not reproduce
9
Misleading notions of "illiquidity risk" and "illiquidity premium" to be given up¡ "Illiquidity risk" refers to variable and unpredictable shifts in tradeability of assets è not
applicable to private markets¡ "Illiquidity premium" refers to reward for investors giving up their ability to freely trade
assets (fixed income) è not applicable to private equity/real assets
Sources of risk are to be found elsewhereØ Measurable¡ Historical losses (frequency and amount)¡ Divergence from average: pooled returns, top/bottom performers, vintage, liquidity
Ø More difficult to assess¡ Concentration and contagion risk (diversification)¡ Behavioural biases (home, prospect theory, overconfidence, timing, fashion/fads,
representativeness, aversion to ambiguity, under-reaction to information, noise)¡ Commitment (intrinsic risk)
Consequences
NRI & eFront – C. Demaria – 19/2/2019Material under copyright – Do not reproduce
What matters: potential losses
Assessing frequency and amounts at stake
11
Losses Profits
Distribution of private equity funds-of-funds based on their IRR
Frequency and losses in private equityUsing funds-of-funds as a proxy for private equity index-fund
N = 346. Source: Wellershoff & Partners/Cambridge Associates, 2017. Data as of Q2/2016.
• Low loss ratio, and limited losses• PE = less risky than listed assets!
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
Strategy- and execution-related risks
The variation of pooled returns and the divergence of individual vintage years
13
Strategy-related risk: a macro aspect
Source: Cambridge Associates, Wellershoff & Partners, 2018.
Evolution of the pooled average TVPI of global small and mid-size LBO (fully realized)
• Significant variation of performance over time (macro-related risk)• No loss on a pooled-average basis (but CA data is skewed upwards)è Important for portfolio construction and scenario modelling
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Multiple (TVPI)
Strategy- related risk
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
14Source: eFront Insight, Wellershoff & Partners, 2018.
Evolution of the TVPI of active LBO funds compared to historical average
Strategy-related risk: individual vintage years
• Tracks active vintage years è benchmarks VY and individual funds• Usable to assess opportunities on the secondary market as wellè Use for tactical purpose and active portfolio management
Historical average
2009
2010
20112012
2013
2014
20152016
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Mutiple (TVPI)
Quarters
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
Selection risk
Extreme and most frequent selection risk
16Source: Cambridge Associates, Wellershoff & Partners, 2018.
Evolution of the top/bottom 5% & 25% TVPI of small and mid-size LBO (fully realized)
• Selection risk is not symmetrical, nor constant• Even worse performers can do well… and are more stable è less « risky »?è Higher performers exhibit a higher dispersion
Assessing selection risk
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Multiple (TVPI)
Top 5% Top 25% Bottom 25% Bottom 5%
Extreme selection risk
Most frequent selection risk
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
17Source: Cambridge Associates, Wellershoff & Partners, 2019.
Correlation of TVPI and TVPI spread (5%) of small and mid-size LBO (1986-2005)
• There is a direct link between TVPI and extreme selection risk• Risk is related to the maturity of funds è limits the assessment of recent fundsè Possible to calculate a simplified/adjusted equivalent of Sharpe/Sortino ratios?
The acid test: risk and return correlation (1/2)
R² = 0.36367 R² = 0.42291
R² = 0.87918
R² = 0.49537
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
0 0.5 1 1.5 2 2.5 3 3.5 4
TVPI
TVPI spread (5%)
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 All Linear (All) Linear (93-99) Linear (00-05) Linear (87-92)
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
18Source: Cambridge Associates, Wellershoff & Partners, 2019.
Correlation of TVPI and TVPI spread (25%) of small and mid-size LBO (1986-2005)
• There is a direct link between TVPI and most frequent selection risk• As time goes by and the industry matures, the relationship increases
The acid test: risk and return correlation (2/2)
R² = 0.18168 R² = 0.53206
R² = 0.92178 R² = 0.06857
1.00
1.20
1.40
1.60
1.80
2.00
2.20
2.40
2.60
2.80
3.00
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
TVPI
TVPI spread (25%)
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Linear (All) Linear (93-99) Linear (00-05) Linear (87-92)
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
What about... liquidity risk?
Liquidity is a dimension of investment: risk can be assessed
20Source: Cambridge Associates, Wellershoff & Partners, 2018.
Correlation of TVPI and time-to-liquidity of small and mid-size LBO (1986-2005)
• No relationship between performance and time-to-liquidityè Relationship independent of the period considered
Liquidity: independent from returns
R² = 0.00635
R² = 0.41819 R² = 0.32093
R² = 0.0079
1.00
1.20
1.40
1.60
1.80
2.00
2.20
2.40
2.60
2.80
3.00
0 1 2 3 4 5 6 7
TVPI
Time-to-liquidity 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Linear (All) Linear (93-99) Linear (00-05) Linear (87-92)
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
21Source: Cambridge Associates, Wellershoff & Partners, 2018.
Correlation of TVPI spread and time-to-liquidity of small/mid-size LBO (1986-2005)
• No relationship between risk and time-to-liquidityè Liquidity is not a risk in itself, but there is a liquidity risk. How to assess it?
Liquidity: independent from risks
R² = 0.01022
R² = 0.16458
R² = 0.07254
R² = 0.02721
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
0 1 2 3 4 5 6 7
TVPI spread (5%)
Time-to-liquidity 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Linear (All) Linear (93-99) Linear (00-05) Linear (87-92)
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
22
0.00
0.50
1.00
1.50
2.00
2.50
3.00
0 1 2 3 4 5 6 7
TVPI
Time-to-liquidity (years)
Min Average Max
VY 1993 3.23 years
Source: Cambridge Associates, Wellershoff & Partners, 2018.
Relationship TVPI and time-to-liquidity of small/mid-size LBO (1986-2005)
• Liquidity risk can be understood as a continuum: 3.23 to 6.48 years• Average is 4.24 years, the pivot for portfolio constructionè Lower and higher liquidity thresholds are the risk to consider
Liquidity: measuring risks
VY 19876.48 years
4.24 years
Risk: shorter durationConsequence: opportunity cost
Risk: longer durationConsequence: payout shortfall
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce
Conclusion
24
Conclusion
Positive features of these instruments¡ They can be applied to listed assets¡ They are compatible with mid- to long-term investment horizons¡ They support the set-up of ratios (Sortino equivalent) with bespoke thresholds (0, minimum
return)
More pratical measures could be identified¡ Concentration risk¡ Pro-cyclicality¡ Under-reaction¡ Vintage auto-correlations
Note: this is a work in progress, product of a book under way
Thank you for your attention!
Wellershoff & Partners– C. Demaria – 30/1/2020Material under copyright – Do not reproduce