Stanford GSB Sloan Study Notes, Week 6 (26), Winter quarter
The picture reflects the exact view I have from where I’m sitting posting this. When you live in Northern Europe, “good snowy ski weather” usually means you have to give in on other things (such as clear skies and light). Not in Tahoe – and that’s why a mission of Sloans have landed here again for a long Presidents’ Day weekend.
But no play without hard work, right. There were quite a bit of extra-curricular activities on campus (I made it to several BBLs even!), some long-planned and inspiring 1:1 coffees with MBA colleagues, a few guests I would bucket in “personal heroes” category. And a fun roleplay of 8am termsheet negotiations, with lawyers at each table and all.
Covered in this issue:
- Differences in financing with debt vs equity – and some irrationalities caused by taxation
- Seed financing – how to survive until Series A
- Intricacies of convertible note structuring
- European startups: plasma drilling in Slovakia and why you should move to Berlin
- When and why founder CEOs get fired
- Introduction to Private Equity
- BBLs on Crowdfunding and Big Data
- More guests from: Geothermal Anywhere, Soundcloud, Intellicap, Twitter, Benchmark Capital, Hellman & Friedman, PubVest, LinkedIn
FINANCE 229: Sloan: Core Finance (Strebulaev)
- Most difficult thing in life is estimating the mean of a random process. (On estimating the forward looking market risk premium). Mathematically, 100 years of data is not enough for either corporate bond yields or average daily temperatures – need around 1200 years of monthly data points for reliable estimates.
- Wason selection task – famous test in the psychology of reasoning
Firms Funding Investments
30% of financing in US companies from debt on average (extremes: consumer banking 85%+, software <5%)
Important video on having too much leverage:
100% debt financing would be virtually the same as 100% equity: there would be no preference order any more
Why investors demand higher premium on leveraged equity? When you do the math and add any debt-based (riskless, predictable return) cash flows to any equity projections, you will drive the expected returns and volatility/risk up for the entire cash flow.
- For the same reason, arithmetically the EPS will be higher for an investment financed using debt (as the number of shares increases when using equity), but so does the risk due to leverage
Dividends are a soft commitment, interest hard – if you don’t pay, you go bankrupt. Traditionally regulated industries issue a lot of debt, because that is the way they can control cash flow back to investors and avoid unions diverting free cash to pay and other cost rises.
Modigliani – Miller (M&M) theorem (2 Nobels 1958). In the absence of imperfections the right hand of the balance sheet doesn’t matter. How we finance investments doesn’t impact the cost of capital (WACC).
- Focuses us to look for imperfections that would render debt or equity more beneficial: taxes, bankruptcy cost, incentives etc
Valuation with Taxes, WACC and APV
- Key assumptions for perfect capital markets: no taxes, bankruptcy costs, agency costs, symmetric information and fair pricing of securities. If the perfect market existed, you would prefer to raise money 100% from debt.
- Corporate income tax in US only since 1909, in UK 1972. In the beginning interest and equity dividends were treated equally, during WWI one of the paragraphs were cancelled, the debt/interest one remained.. Only one country in OECD still allows tax deductions on dividends: Belgium
- In US, dividend yields have consistently dropped over last 40 years, largely because of the tax disbenefits
- Equity holders have the residual claim on corporation’s cash flows (because of the preference of debt). Residue is more volatile.
- V (Levered Firm) = V (Unlevered Firm) + PV (Tax Shields). After some simplifications: PV(Tax Shields) = Corproate Tax Rate * Debt. E.g, with 35% tax rate in US, issuing $1B of extra debt increases your equity value automatically by $350M.
- Leveraging recapitalization (levcap): using debt to buy back your own shares.
- Adjusting WACC to taxes: just multiply the debt component of cost of capital with (1 – Corporate Tax Rate). But there are limits, if there are too many and too large tax components, WACC is not the best method any more, should prefer Adjusted Present Value (APV).
- If you apply both correctly, you get the same answer – but gets harder to do WACC correctly.
- WACC requires a constant debt ratio in the future (constant debt level, full payoff of principal in 5 years, etc don’t fit the model)
- WACC also becomes useless with very high debt ratios (like leveraged buyouts)
FINANCE 373: Entrepreneurial Finance (Korteweg)
- Convertible notes becoming more common (2011: 41% of seed rounds VS 59% preferred stock)
- Conversion based on A round price is automatic under minimum size of A round. Discount, cap and other features are optional. Common lengths today 12-24 months, on 6-8% interest.
- Cap is usually set as fully diluted (this early usually founder stock only), pre-money valuation of A-round. The A-round investors don’t use the cap as an input when valuing the round, but it could have indirect implications (psychological anchoring, different investor incentives)
- Angels, incubators care about the cap because they are likely not going to invest in future rounds (and would get diluted aggressively). VCs care less, because the seed rounds are relatively tiny compared to anything in the future.
- Series A investor gets their intended share (say 40%) of the share in company as it is after founder stock plus converted seed investors.
- Bootstrapping entrepreneurs usually do a regular note with themselves (not convertible). VCs are fine with that because they are still negotiating with a single person (the founder) – if there are outstanding notes with other parties, VC will ask you to go an renegotiate those.
- Angel VS VC agency conflict: first are happy with much smaller and earlier exits.
- Decided to raise seed money based on the team. Even delayed building out the prototypes a bit and pitched based on Powerpoint – to make sure investors invest for the right reasons.
- If you have a VC in seed round and for some reason they don’t follow on in the A-round – you’re toast, no one else will either. Puts some extra pressure on founders.
- Be upfront with angels who are worried about dilution: if you’re building something where you’re expecting to need to raise tens of millions – they will get diluted heavily and should only join if they are comfortable with that.
- Typical seed round milestones: hire the team, end user traction, premium experiments.
- Test the equity setup on the market. Got proposals to do a priced round and richer terms from elsewhere (10M+ caps), but the majority of market seemed to talk convertible notes and we came back to the investors we wanted to work with.
- Yousendit took 7-8 months to raise $50k even with millions of users already. It is all about reputation and track record. Very hard if it is your first time.
- No science in cap setting. If a Y-Combinator company sets $5M cap, everyone has $5M for a while. If there is a runaway with $10M, everything in the same class goes up for a while.
- These days a few TechCrunch articles and going to This Week in Startups and you start getting incoming calls “can we see your stats, are you thinking of A-round?”. Yet, in a hot market you probably need to turn some amazing terms down because it is too early – and if you ever want to have a chance to raise an up round.
- Reasonable angel ask: hey, we are in for the round, but please go visit these of our VC friends too to give them an early look. We want some feedback for you from the people we’re feeding into.
- “Add water” type of business: just add more sales people and it scales.
- Founder should be a little selfish, build a company in a way that helps you scale, not just the business.
Case: eDocs (termsheet negotiation role play)
- benchmarking valuation against anything else than VC term sheets (like “we could seek the entire company for $15 now – why are you offering $6M?!”) sends a lot of bad signals: if you would honestly consider cashing in right now, why would we believe you’re in it for $200M future in 5 years?
- you can use warrants to adjust valuation if the VC+Entrepreneur can’t find a planned second VC to come in: investing $2M at $1M, trying to find another $2M – if you can’t, investing $4M alone, but get 0.5M warrants for extra shares at $.10 (total valuation goes down)
- if you leave founders (or common stock) board seats after they leave/terminate, you are potentially seeding very bad ongoing tensions on board level, should their departure be messy
- double trigger vesting with 100% acceleration is currently a market standard
- non-partipating preferred stock is actually good for the longer term future of earliest VC (can sound counterintuitive) – because if that is in Series A, will be inherited by all later rounds with higher preference
- voting setup with 2 VCs: 60% vote to get stuff done, 50% veto rights: dives preferred stock consensus
ME421: European Entrepreneurship & Innovation Thought Leaders (Lee)
Slides and videos: http://www.europeanentrepreneursatstanford.com/
- Geothermal energy is currently feasible only at 3-4-5km depths – could be used anywhere in the world if it was reasonable to drill 10km
- Geothermal Anywhere has built a plasma drilling product called PlasmaBit
- Some current processes are not even feasible: for example fracking for shale gas in Nigeria can not be done with today’s tech because a barrel of water is more expensive than barrel of oil
- Started 2007, about 170 people now (110 joined last year) in Berlin, London, Bulgaria, San Francisco
- A-round in 2009 was turned down by everyone (mostly in London and US) – Sequoia had just sent out their “RIP” deck, music space was toxic, etc…
- Working on a book on social web in 2006 while having an issue sharing audio files for comments from friends as a sound engineer
- Staying in Europe as a way to stay closer to users, as opposed to getting too close to technologists. Decided in the evening of one-day visit to Berlin that they will move from Stockholm there next week, because of the good vibe. Intersection of technology and art, a city with punk ethos.
- Berlin is becoming a default hub for design heavy, consumer internet startups built by expats from elsewhere in Europe. Life happens in East Berlin: MItte & Kreutzberg
- The same Swedish characteristics that lead to great Scandinavian design (desire to not stand out) can be counter-effective for entrepreneurship
- Cofounder dynamics: chances of success are higher for the company if at least one of them is not having a bad day and can cheer others up
- A good book on deriving incentives from Maslow pyramid: http://www.amazon.com/Peak-Great-Companies-Their-Maslow/dp/0787988618
- Experienced people tend to have more roots around them, so it is harder to relocate them
- The worst part of the Valley talent wars is the distraction it creates: you know that the best engineer you hired today gets attacked 15 times a day. The best startup engineers today are in Eastern Europe.
- Sound is going to be bigger on the web than video – so it is about bumping YouTube from #3 to #4.
- Podcast is a delivery mechanism, doesn’t have anything to do with content really. Imagine you could use YouTube the same way: subscribe, download every video and then watch…
FINANCE 385: Angel & VC Investments (Strebulaev)
Guest: Vineet Rai (Chairman & Founder of Intellicap, India)
- Started with $2100 11 years ago, now a $110M fund. $20k-$3.5M investments, almost always the first investor.
- Invested in 60 companies, abnormally high success rate for pre-revenue seed investments (has lost only about ~10 co-s)
- Underserved, high risk market is a huge opportunity for venture capital, with additional motivation of spill-ove social good of entrepreneurship
- “I was a good guy investing in good guys – but it turned out to be a bad business.” Later came out it is important to find good people who can manage scale – a good doctor might not necessarily be able to run 50 hospitals.
- Good entrepreneur does everything for the cause of their business, not necessarily the cause of his investors.
Contracting Between Angels/VC & Entrepreneurs
- Standard difference of debt over equity are seniority, interest obligation and no right to vote. Convertible notes for startups are not standard debt, as they often behave more similarly to equity.
- Since 2010 two trends combining: a) many more seed funds including VCs and b) entrepreneurs seeing value in not pricing the round yet (and thus offering convertible notes)
- Convertible note legal language is still not as stable as VC term sheets, some examples:
- Consideration: amount invested in company
- Principal: amount of money investors are owed by firm
- Unpaid accrued interest: as startups don’t have cash, interest accrues towards the round without getting paid
- Maturity: time when debt is due – really a deadline by which more money needs to be raised (to get a real valuation). Often renegotiated when arrives before conversion – as there is no cash to be returned anyway.
- Seniority: Convertible debt is senior to convertible stock
- Outstanding amount of note: principal + unpaid interest
- Prepayment: a possibility for company to return the Outstanding amount early (usually requires agreement from majority of shareholders)
- Conversion premium: fancy way to say “discount”
- Typical conversion rules:
- Outstanding amount is converted…
- …at a pre-defined rate (with or without discount)…
- …at the time of A-round…
- …into (virtually) same security as issued to Series A investors (the exception being liquidation preference)…
- …with automatic conversion (conversion is not optional).
- Discounts compensate angels for both investing into a riskier company and time value for earlier investment. The common 20% is heavily debated in early investment circles, if it compensates angels fairly enough for both.
- With increasing Series A valuations, angels’ fraction of the company is reduced. Their shares total value equals
- (Consideration) * (1 + Interest) / (1 – Discount)
- As the Stock Price is cancelled out above and below the fraction, this formula is the same regardless of the next round valuation!
- This makes uncapped notes very similar to vanilla preferred stock (seniority, but no real participation in the upside, increasing dilution with higher valuation rounds) and the relative return on investment is mostly dependent on the discount.
- Solution: setting a predetermined cap. A hint of valuation without setting a valuation.
- (Consideration) * (1 + Interest) / (1 – Discount)
- Interesting dynamic: 2010->2011, share of closed deals with cap has remained the same ~82-83%, BUT the median valuation cap has grown $4M -> $7.5M! (Fenwick & West Seed Financing Survey)
- Conversion when hitting the cap very simple:
- (Conversion price for angels) = (Cap) / (Number of Common Stock Shares)
- VCs are incentivised to renegotiate this if as a result of the conversion the angels go into a territory (say 40%) that would make the math impossible for the VC to get ~30% of the Series A round and still have motivated founders. But this is a once in a decade scenario.
- If you receive two, otherwise identical offers for $0.5M investment: 1) $0.5M convertible note, no discount or interest, with $5M cap or 2) $0.5M convertible preferred stock with $5M pre-money seed valuation. Which one do you choose, expecting a $10M+ Series A and an exit even higher?
- In optimistic scenario, the payoffs are identical
- In down round, however, the convertible note holder would be better off (and entrepreneur worse) because of the seniority of debt. Yet, having a Series A round below seed round cap is virtually impossible (who would buy into a first round of something that is already failing?) so you would have failed then anyway.
- In reality, convertible note vs preferred stock offers are almost never equal for other things (boards seats, voting rights, etc) – VCs (preferred stock) usually request much more than angels.
- Convertible notes, due to lack of valuation round, do not start the US “tax clock”
STRAMGT 354: Entrepreneurship & Venture Capital (Wendell)
Unfortunately can not publish my notes from (fantastic) class discussions – there is a no-blogging policy to protect honest conversations and especially the guests.
- Changing the IRR expectation for the deal you’re valuing is a very blunt mental instrument, that basically does a haircut across the board. Better tool is to actually go into the details of the business plan, adjust per-product estimates and timings. Among other things this will increase your understanding of the business and creates room for actual dialogue with the entrepreneur.
- Early stage CEO should assume they can not indefinitely maintain secrecy about how much stock people have. A benchmark if the structure is fair: should the list of grants become public, would the CEO be able to convincingly defend the distribution? Doesn’t imply equality, but logic.
- Replacing the CEO when the product/sales are picking up might seem extra unfair, but is usually good for the company. Can bring new one in easier, people believe in the future potential, there is more inertia to growth through the disruption of change.
- Reed’s blog post on the founders & CEOs is the most enlightened on the topic.
Article: Venture Capitalist and CEO Dismissal – a paper examining the actions a VC desires of the CEO, reasons CEOs are replaced and the impact of replacement to ventures performance
- A founder works with limited resources, gets hands on and hires for the competence gaps. A CEO needs to also hire for things they could do, let go and delegate to scale. Therefore, logically replacing a founder who is not scaling will leave the largest holes in the company (because he didn’t delegate)
- There is a big bias for pattern recognition, and development from founder to CEO is a lot about breaking out of the patterns.
- Change in qualities and techniques needed for running a tech company is a highly correlated to the number of staff.
- Relative importance of new venture CEO activities (survey of US VCs). Top answers on scale of 1 (unimportant) to 7 (very important), all with std dev <1:
- Strategic Decisions (6.60)
- Staffing Firm (6.47)
- Strategic Planning (6.40)
- Motivating employees (6.36)
- Dealing with Crisis (6.36)
- Allocating Resources (6.19)
- Working with Board (6.19)
- Managing Growth (6.00)
- Same respondents asked to identify up to three activities where CEO failure lead to their dismissal (% share of all reasons cited):
- Making Strategic Decisions (16%)
- Working with Board (14%)
- Motivating Employees (12%)
- Allocating Resources (12%)
- Strategic Planning (10%)
- Staffing Firm (9%)
- Managing Growth (8%)
- Dealing with Crisis (7%)
- Never cited as cause: Ceremonial Responsibilities, Disseminating Information, Negotiating
FIN587: Private Equity (Parker)
- Prof Parker is GSB’62 – the term private equity didn’t exist
- Private equity was born to solve the problem where many companies were mismanaged, management resisting to change and no single shareholder having enough clout to make them change
- Only four primary categories of management PE can use to create or “liberate” value. Note that EVERYTHING could be done by the target company by themselves:
- distributing excess cash to owners
- eliminating wasteful expenses (trimming the fat, not muscle) – that is too painful to do from inside
- increase leverage (borrowing in lieu of equity) – debt can be dangerous and risky, but the discomfort is what generates returns
- selling or spinning off assets – no nostalgia. finance, not romance.
- (could argue strategic input & guidance is a separate, 5th one?)
- Exits: IPO, resale already public shares, mergers, breakups to sell piecemeal
- Management incentives based on rewards (pay) and punishment (getting fired). PE is just more extreme on both dimensions: pay better & fire quicker. Why don’t the managers have the right incentives, because the boards are too much club like, and not really motivated from value creation upside (vs PE 20% carry).
- Misleading to think of PE as an asset class – it is rather a governance model.
- Stanford endowment has $10B of illiquid private equity holdings.
- If a PE even goes in as minority investor, they will definitely be on the board and a loud minority.
- Industry dynamic – 20-30-40 years old firms rarely have their favourite deal from last 5 years. Usually there is something transformational earlier on that makes the firm.
- Hellman was also the founder of Matrix Partners (VC)
- 26 years, $16.6B invested in 78 companies. Fund VII currently investing $8.9B
- 45 people in investment team, 30 working on deals (3-4 months work per one deal, 20% close). $800M of GPs own money along with LP-s
- H&F is based on book value: passed on from generation of partners to another on book value, not market value. Enforces real value generation.
- One more value creation opportunity: identifying and pushing growth initiatives, which the management has become to risk averse to launch.
- Have return strategy, not growth strategy.
- Strong in large-cap software companies.
- With 1-4 investments a year you can not get to quantitative evaluating your people, for example. A lot of qualitative assessment, but pattern recognition still develops over the years.
- Ongoing tracking of final investment committee case: people on the deal need to report back against it every 6 months.
- Investment committee decides both entry and exit. Small committee (just 4 people) assure full accountability, their careers are on the line – can’t just provide “casual opinions” as when there was a 10 person committee. Also, small teams can realistically operate on consensus, not split voting.
- Highly centralised as we do investments, highly decentralised on how people choose to spend their time.
- Average tenure 10 years for Managing Directors, 18 years for IC
- We don’t do hostile deals. Ultimately all deals are friendly, you sign it, you have a party – ok, well, not everyone might like each-other at the party…
- People generally know the value of their companies. You rarely “get a steal”, need to have a differentiating point of view, value creation idea, better access / relationships to get the last look
- DoubleClick case: auction to take it private, 70 offers invited. 3 person deal team did nothing but this for 4 months.
- Senior management replacement recruitment often starts in parallel to due diligence.
- If you pay the tuition (in form of lost money in underperforming business), you better take the time to absorb the lesson.
- Like the companies, not industries.
- F1 was bought and sold in 30 days
- You think about the exit strategy up front, but you don’t define it and don’t spend too much time on it. More general belief that there are always people who want to buy a great company, given you do a good job turning it around.
- Pre-MBA: associates, post-MBA: directors/principals
- Working in PE gives an opportunity to work with very high potential companies, while still being in a small organisation, with a lot of influence from very early on.
BBL: Crowdfunding 101
Guest: Brian Goldsmith (PubVest)
- Since April 2012, JOBS Act allows small business to raise money for their securities over the internet. Not fully operational yet, SEC is writing their detailed rules.
- Before: only 8.3 million Americans qualify to be accredited investors (>$200k annual income or >$1M of assets beyond primary residence) and still needed to be acquainted with the entrepreneur to legally invest.
- JOBS Act allows
- non-accredited investors can now invest 10% or $100k of income over 100k income, or 5% or $2k if under $100k
- simplifies for accredited ones (no registration necessary)
- Income, or especially inherited assets really don’t indicate for investor sophistication (is a Stanford PhD in engineering less skilled to invest in a tech startup than your rich aunt? Singapore has opted to use real testing of investment competencies instead.
- Worldwide crowd funding from 2009 $0.5B -> $0.85B -> $1.5B -> $2.8B in 2012
- PubVest is a curated marketplace that screens companies raising money and lives off fundraising success fees
- A notion of a funding portal, that can do much less than a current broker-dealer can. Can share investment opportunities, but can not hold investor funds, give advice, pay its employees for performance, etc. Because it is so limited, PubVest is actually a broker-dealer.
- AngelList is using an exception in the current law called Private placement. You can “like” a company on AngelList, but when you have self-sertified as accredited investor, you are diverted to another web page that talks about a private placement to another entity than the one was originally listed. Private placements are void if they become public. Complex to manage.
BBL: Big Data, Deep Insights, Fast Actions
- LinkedIn: 200M users (+2/sec), 2.6M+ company pages, 4.2B searches in 2011
- Business analytics support 70% of 3500+ internal employees
- Classifies data into four, based on volume & analytical challenge/complexity: ERP data, CRM data (gigabytes), web data (terabytes), social data (approaching infinity on both axis)
- imaginary user with only social data (not active on web, no CRM/ERP data) can still be profiled by guesstimating his attributes from those of his social links
- LinkedIn data:
- Volume: Terabytes, records, transactions, tables, files
- Velocity: Batch, Near Time, Real Time, Streams
- Variety: Structured, Unstructured, Semistructured…
- Big Data: not the amount of bytes in your data, but the amount of data in your bytes
- Increase in business value AND analytical challenge: what happened -> why it happened -> what is happening -> what will happen -> can we make it better?
- Ad-hoc analysis should be below (closer to data management) BI & Reporting in data manipulation architecture.
- “Interesting” doesn’t mean anything on data insights, always ask “so what?”
- A data scientists solving 500 requests a year can seem like a miracle worker. Yet, if he is supporting 250 people, every employee gets only 2 requests answered for the entire year!
- Data dimensions: behavioural (logins, site activity…) + identity (country, skills, seniority…) + social (connections & clusters)
- Machine learning algos scanning profiles for full text keywords
- What is the difference between Chopin’s music and email cadence? On some level, it is just the right thing at the same time that people like…
- Web tracking of all events (including many hovers!)
- Apache Kafka for near-time data streaming
- Future challenges: how can we use LinkedIn data to make people more successful? A person joining the right company in the right role at the right time to build towards their career goals?
- Internal “talent flow” experiments: visualizing how good people move between companies
For more posts on the Stanford GSB Sloan life – see the table of contents here.