The Use of Monte Carlo Simulation for Valuations and Return on Investment Analysis in Private Markets

Palisade Annual Risk Conference San Antonio, Texas 12, 13 November 2019

Robert F. Brammer, Ph.D. President and CEO Brammer Technology About Me

› Robert F. Brammer, Ph.D. President and CEO, Brammer Technology › Consultancy in information technology, weather and climate, and financial services – clients include start-ups, VCs, and Fortune 50-size corporations › Brammer – manage 25 investment portfolios for family members

› Formerly VP and CTO Northrop Grumman Information Systems Sector

› Board membership for both public and private corporations

› Participant in ~20 corporate transactions (e.g., M&A on both buy-side and sell-side)

› 25+ year customer for Palisade

› Many years building simulation models using Palisade products– R&D management, electric power trading, cybersecurity, financial analytics, family office operations, economic impacts of climate change, … Key Points for This Presentation

› Private markets, notably and , are growing rapidly and increasingly important in global economies

› Investment returns in private markets are much less predictable than those in public markets, so effective and risk management techniques are critical

› Many possible valuation approaches for individual companies and investment portfolios

› Monte Carlo techniques can strengthen valuation approaches and risk management

› We will discuss the use of the Palisade DTS for model development, equity valuation, and portfolio management

› This presentation is for education about the use of @Risk and not investment advice Organization of this Presentation

Introduction – a few points about private markets

Valuing a company from seed stage through exit

Return on investment for different investor classes

Mitigating investment risk in a private market portfolio

Concluding Remarks and Q&A Introduction Some Trends in Private Markets

› Private markets include corporate M&A, private equity, venture capital, and other deal types (e.g., real estate, private debt, infrastructure, etc.) › We will focus on equity valuations in this presentation Data from Pitchbook Debate over the Decline in the Numbers of Public Companies

› The number of US public companies has declined in recent years

› So what? Some Trends in Public Markets

Number of US public firms dropping Mkt Cap of US public firms as a % of US GDP

› Due to increases in M&A and LBOs, the number of US public firms is dropping › Increases in PE and VC funding is enabling companies to stay private longer › Although public markets are growing as % of GDP, there is an increasing need for analysis of private companies › Financial information service companies expanding into private market coverage (e.g., Morningstar bought Pitchbook) › Investors and portfolio managers want viable choices Data from World Bank Private Market Returns Much More Variable Than Public Market Returns

› 5-year annual returns from US private equity funds and US mutual funds by performance percentile, 2013–2018 › Data from McKinsey Global Private Markets Review 2019 Valuing a Company from Seed-Stage Through Exit Equity Valuation in Private Markets

› Five classes of valuation techniques commonly used › Three market-based approaches › Multiples – e.g., P/E, P/EBITDA,… › Industry valuation techniques, e.g., comparison with public firm valuations › Available Market Prices – e.g., prior transactions involving “similar” private firms › Income approach – DCF › Replacement cost – net asset evaluations (e.g., Intellectual property portfolios)

› Appropriate technique depends on company status, market conditions, … Stage Definitions and Applicable Valuation Approaches

› Angel or Seed Rounds: - a small round designed to get a new company off the ground. › Investors generally include individual angel investors, investor groups, friends, and family. › Valuation methods are very informal, often based on money needed to get a prototype built

› Early stage Rounds -- investment from a VC firm and includes Series A, Series B rounds › Range on average between $1M–$30M. › Valuations often based on ratios and comparisons

› Late Stage Rounds -- Series C rounds and onwards are usually $10M+ and often much larger. › Valuations often based on financial performance with DCF techniques. However, many examples of “story ” with no revenue – more common in biotech and pharma

› Exit Rounds – IPO, M&A, LBO, …

› Further information about other financing methods available from Pitchbook, Crunchbase, and many other sources Open Low High Close Comps 20 20 25 25 Precedents 23 23 31 31 DCF - base case 18 18 24 24 Use ofDCF Multiple - blue sky Methods26 in26 Company32 Valuation32 52 wk hi/lo 16 16 22 22

Valuation Summary $35 $31 $32 $30 $25 $24 $25 $22 $26 $20 $23 $20 $15 $18 $16 $10 Comps Precedents DCF - base case DCF - blue sky 52 wk hi/lo

› Most firms use multiple methods for valuation negotiations

› The visualization is called a “football field diagram”

Data from the Institutee Some Early Stage Valuations

R2=0.58 P<0.0001

› Comparing first deal size to first valuation for 225 start-ups from all 11 GICS sectors that later had IPO events › Deal size - $16M (mean), $6.3M (median), $417M (max), $0.02M (min) › Post-Valuations - $33M (mean), $18M (median), $700M (max), $0.08M (min) › Data typical of private market data with a cluster and a few extreme outliers

› Used @Risk distribution fitting function to the ratio of 1st valuation/1st deal size › Given the importance of ratios in early stage valuations, a distribution provides useful market insight › Model predicts, for example, the ratio has a 50% probability of being between 1.68 and 3.89 – useful for business planning › @Risk selected inverse Gauss distribution (#1 – AIC, BIC, K-S, A-D #4 – Chi-square) Data from Pitchbook Valuation (DCF) Processes

› DCF valuations generally start with a financial model of the company (late stage) › Integrated income statement, balance sheet, cash flow statement, various supporting schedules, … › Model provides annual estimates of

› Used variable cashflow growth rate based on model fit to aggregate annual S&P 500 cashflow per share growth › Mean – 12%, median – 3%

› Used two-phase DCF with termination model with model in place of static growth rates › How does share price grow using sum of discounted cash flows for ten years (reduced growth in 6-10) › Model shows 80% probability of share price increase over a ten-year period, 35% probability of more than 2X increase Data from Bloomberg Comparing the IPO Offer Size to the Total Funding Raised

› Typical IPO data behavior – a few significant outliers and a bulk group Data from Pitchbook › Many offer sizes are less than total capital raised during pre-IPO stages › Filtering data will change resulting model – useful for analyzing prospective IPO price What Happens on the First Trading Day?

› The offer price per share is usually announced the day before the first day of trading

› The open price is the price per share in the first transaction

› Data here for selected US IPO’s in the past few years (120 companies) › Uber – offer $45, open $42 › Lyft – offer $72, open $87 › Ultragenyx – offer $21, open $45.80

› Mean return – 23%, median – 17% › Probability open > offer - ~80%

Data from Bloomberg Return on Investment for Different Investor Classes Cap Table and Return on Investment Example - 1

› The capitalization table (cap table) shows the details of company ownership at a given time › Investor class, , participation rights, and other factors

› At liquidation (e.g., acquisition), the cap table determines the distribution of the proceeds

› A very simple example for illustration › Series B - $7.5M, 40% ownership, liquidation preference - $7.5M, participation cap 3X › Series A - $5M, 25% ownership, liquidation preference $5M, participation cap 2X › Common shares – 35% ownership › Stacked preferences At exit, who gets what? Cap Table and Return on Investment Example – Part 2

› Median - $12.9M, Mean - $19.6M › Prob. of <$7.5M~ 22.2% Prob of < $12.5M = 48.4%, Prob of > $23M = 25% Cap Table and Return on Investment Example – Part 3

› Series B investors receive mean value of $10.8M, median value of $7.7M › 22.2% probability of < $7.5M invested, 25% probability of > $11.6M (55% gain) Cap Table and Return on Investment Example – Part 4

› Series A investors receive a mean value of $6.1M, median value of $5M › 22.2% probability of zero return, 48.4% probability of loss, 25% probability of return > $7.6M (52% gain) Cap Table and Return on Investment Example – Part 5

› Common option investors receive a mean value of $3.6M, median of $128K › 48.5% probability of zero value, 19.4% probability of proceeds > $5M Sensitivity of Investor Procedings to Investment Terms

Case 1 Case 2 › Investments › Investments › Series B - $7.5M, 40% ownership, liquidation › Series B - $7.5M, 20% ownership, liquidation preference - $7.5M, participation cap 2X preference - $7.5M, participation cap 2X › Series A - $5M, 25% ownership, liquidation › Series A - $5M, 15% ownership, liquidation preference $5M, participation cap 3X preference $5M, participation cap 3X › Common shares – 35% ownership › Common shares – 65% ownership › Stacked preferences › Stacked preferences

› Proceedings › Proceedings › Series B - mean return of $10.8M, median return of › Series B - mean return of $9.2M, median return $7.7M, 22.2% probability of < $7.5M invested, 25% of $7.6M, 40% probability of < $7.5M invested, probability of > $11.6M (55% gain) 25% probability of > $9.6M (28% gain) › Series A - mean return of $6.1M, median return of › Series A - mean return of $4.7M, median return $5M, 22.2% probability of zero return, 48.4% of $5.1M, 22.2% probability of zero return, probability of loss, 25% probability of return > 48.4% probability of loss, 25% probability of $7.6M (52% gain) return > $6.6M (31% gain) › Common - mean proceeds of $3.6M, median of › Common - mean proceeds of $5.8M, median of $128K, 48.5% probability of zero proceeds, 19.4% $205K, 48.5% probability of zero proceeds, probability of proceeds > $5M 28.9% probability of proceeds > $5M Mitigating Investment Risk in Private Market Portfolios Portfolio Management in Private Markets

› Private market investment firms tend “Top performers are considering to focus on certain market sectors adjacencies that are one step removed › Staff expertise, social networks, etc. from the core, rather than two or three steps removed. The best options take › However, they do look for some advantage of existing platforms, investment themes and expertise” diversification in strategies, geographies, etc. to balance risks in GLOBAL PRIVATE EQUITY REPORT 2019 investment portfolios Bain and Company › Implies different return models and some levels of correlation

› The following example uses RiskOptimizer to shape the distribution for a portfolio of ten companies with modeled distributions and share price correlations Portfolio Optimization Using RiskOptimizer

Company Capital Raised ($M) Shares (M) Model Ent Value ($M) Share Price ($) Min Cost Max Shares Purchased (M) Port 1 Value ($M) Co.1 100 3 1.18 $118.33 $39.44 $0.00 $2.00 $10.00 0.073 $2.89 Co.2 80 3.5 2.63 $210.67 $60.19 $0.00 $2.00 $10.00 0.073 $4.42 Co.3 65 5 2.52 $163.58 $32.72 $0.00 $2.00 $10.00 0.073 $2.40 Co.4 120 8 2.10 $252.00 $31.50 $0.00 $2.00 $10.00 0.073 $2.31 Co.5 30 4 1.68 $50.25 $12.56 $0.00 $2.00 $10.00 0.073 $0.92 Co.6 50 7 2.13 $106.67 $15.24 $0.00 $2.00 $10.00 0.073 $1.12 Co.7 25 4 2.15 $53.75 $13.44 $0.00 $2.00 $10.00 0.073 $0.99 Co.8 50 4 2.87 $143.33 $35.83 $0.00 $2.00 $10.00 0.073 $2.63 Co.9 38 5 2.05 $77.90 $15.58 $0.00 $2.00 $10.00 0.073 $1.14 Co.10 45 3 1.16 $52.13 $17.38 $0.00 $2.00 $10.00 0.073 $1.27

Total Cost $20.00 Total Portfolio Value $20.09 › Input to RiskOptimizer includes a ten-equity portfolio with a model for the distribution of the enterprise value for each company and uses the number of purchased shares for each company as adjustable cells

› RiskOptimizer enables portfolio shaping with a variety of optimizing statistics, including mean, median, variance, percentiles, …

› If all else fails, use the “spray and pray strategy” Concluding Remarks Concluding Remarks

› The growth of private markets leads to increasing importance for valuation

› In many cases, data sets for private companies are incomplete as compared with data sets for public companies › Implies needs for modeling and analytics

› Modeling pre-IPO companies can be particularly difficult due to presence of extreme outliers

› Be very careful with VC termsheets

› The results shown here potential value of further related research on private markets using data fitting and Monte Carlo simulation

› I am looking forward to seeing how DTS 8 can help. Robert F. Brammer, Ph.D. Brammer Technology President and CEO Andover MA 01810 Brammer Technology [email protected]