Sean Park Portrait
Quote of The Day Title
Don't look for solutions, look for problems. Look for stuff that seems broken.
- Paul Graham, Y Combinator (on how to find ideas for start-ups)

The case for investing in new companies.

Buttonwood has posted an excellent analysis of why financial markets are unlike other markets for goods and services:

This apparent contradiction can be resolved. Financial markets do not operate in the same way as those for other goods and services. When the price of a television set or software package goes up, demand for it generally falls. When the price of a financial asset rises, demand generally increases.

Which explains why bubbles develop and burst and why ‘market fundamentalism’ does not generally serve us well when thinking about financial markets (as opposed to other markets.)  Buttonwood also alludes to the fact that bubbles often develop at times of great change (has he read Perez???):

Why not just let the markets rip? Some would say that bubbles tend to coincide with periods of great economic change, such as the development of the railways or the internet. Individual speculators may lose from the resulting busts but society gains from their overoptimistic investments. However, this argument is harder to sustain after the recent bubble in which society “gained” some empty condos in Miami and holiday homes in Spain.

His conclusion is that because of these structural characteristics of financial markets, central banks (and possibly regulators and/or governments) have a natural, pro-active role to play in trying to mitigate or counter these problems.

Of course a few investors – the most high profile being Warren Buffet – have successfully arbitraged this weakness in capital markets buy being countercyclical, being “greedy when others are fearful, and fearful when others are greedy”; but as most people know this is bloody hard to pull off and exposes the investor to significant liquidity/solvency risks if they get the timing wrong.   As Keynes said, “the markets can stay irrational, longer than you can stay solvent…”  If you have an edge, even a small one, doubling down will usually work as long as you have an infinite bankroll. Ooops, small fly in the ointment.  (Besides, if you have an infinite bankroll, what the hell do you need to bother about worrying about returns!)

Well I have neither an infinite bankroll nor the skills (and/or luck) to adopt a Buffet-esque investment strategy.  But I do have some skills.  And some experience.  And I can recognise patterns reasonably well.  And I have conviction.  And a reasonable track record for building new markets and adopting and executing novel business models.  So a few years ago I figured out that by focusing these modest talents and skills on investing in and helping to build new businesses, with a lot of hard work and days and months of research and reading I could generate pretty decent financial returns that were (almost) completely uncorrelated with the massive tides that buffet the world’s financial markets.  And most importantly, this lack of correlation is structural – ie it doesn’t disappear in violent bear markets when almost all mainstream asset classes discontinuously jump to near perfect correlation (much to the chagrin of the VaR boys.)

It’s not hard to understand why.  In fact it’s pretty obvious.  For a new business, the ups and downs of the market, GDP, etc. have at best a second or third order effect on the company’s value.  These factors are overwhelmed by the single most important factor driving value creation which is of course, can the company successfully sell it’s products or services to paying customers (or be more and more clearly on that path.)  As someone wise once said:  a “start-up is not GM”  ie They are not correlated to GDP.

Now don’t get me wrong, I’m not suggesting that investing in new companies is without risk.  In fact as most people would glibly observe, investing in start-ups is ‘very risky’.  Well yes, but the risk is almost entirely idiosyncratic and manageable – much much less dependent on vast, uncontrollable, macro-economic trends and forces.  And just because the risks are easier to identify and name, doesn’t mean it is easy to manage them, just that they are potentially (more) manageable.

So if this is true, why have venture capital returns generally been so poor (at least in the last decade or so) and why don’t more smart people try their hand at this (rather than trading/managing other types of assets)?  Answering the second question first, I suspect this is because failing together is much nicer than failing alone, and so if the global financial crisis wipes out your hedge fund or investment bank or savings, well that sucks but, you know, shit happens.  If however you pour your own (or worse your investors’) capital into a couple of dozen new companies that crash and burn, well that’s just a very lonely place to be.  The answer to the first is not simple and you could probably write a book on this (perhaps Paul Kedrosky will?) but with the disclaimer that I don’t pretend to really know, my short and dirty take would be that there are two related factors at the heart of this failure.  First, investing in new companies is hard to scale – at least compared to many/most other asset classes and secondly, the traditional structure of the industry is poorly adapted to this reality.  Private equity legal and economic structures (which is how most venture partnerships are structured) doesn’t really fit the risk/reward/resource profile needed to invest successfully in new companies.  Of course their are exceptions – both temporal and human – but just because their are some investors clever and/or lucky enough to make it work doesn’t make it right.

I could of course be wrong.  And I could fairly be accused of hubris, especially as at this point I don’t have a long enough track record and/or enough exits to prove without doubt that my approach is correct.  And while I am confident in my own abilities and have backed that up with a lot of “skin in the game”,  I am even more confident in my larger analysis that while the venture capital industry might be broken / poorly organized, the risk-adjusted returns available to those who chose to invest – methodically and with a well-calibrated capital and incentive structure  - in new companies, are excellent and, for the VaR-boys out there, truly uncorrelated to mainstream asset classes.  The challenge is of course to find these investors and not to swamp them with too much capital.  This problem isn’t solved but it looks a hell of a lot like the problem facing hedge fund investors (in most strategies that also do not scale beyond certain amounts of capital) and the asset allocation community would do well to try some of their more successful solution there on finding and seeding managers in this asset class.

And if you ask me, the rise of the ‘super-angel’ much talked about in venture circles these past months, is a step in the right direction and perhaps an indication that asset allocators are (finally) waking up to this opportunity.

Related articles by Zemanta
Enhanced by Zemanta

Post to Twitter Tweet This Post

Permalink||Comments|Filed under 'Capital Structure, Private Equity & Venture Capital, Risk Management'
  • Buttonwood's analysis doesn't seem that good to me, and could easily be written off as yet another piece of glib journalism. There's no great new economic insight here, and there's really no difference between a 'financial asset' (which I take to be assets traded on financial markets) and a plain old asset. Rivalrous goods, like land in Mayfair, Ferrari F50s and Berkshire Hathaway stock will always behave differently. iPhones may still be rivalrous in nature, but simultaneous consumption is easier when there are millions of them, and fast depreciation/obsolescence is factored into those markets (the nearest analogue might be clothing - high fashion today, end of line sale next quarter).

    What's special about financial services (and the assets they trade), and what Buttonwood fails to explain, is the fee/transaction cost structure.

    Early stage startups as an asset class are interesting precisely because they haven't yet been fully ensnared into the mass market financial services machine. This is the same reason why hedge funds were interesting a decade ago - before too much capital was committed to them and average returns rapidly converged on those from more traditional fund management structures. It's even possible that seed stage investing might escape from ever becoming mass market, constrained as it is by the lack of quality entrepreneurs and the need for subject matter expertise.
  • I remember there being a graph showing that US households overall exiting the NASDAQ during the boom, and were replaced by funds & pensions. A lot pro-cylicality seems to come from 'I'll be gone, you'll be gone.'
  • Hi Sean,

    I have 2 comments:

    First, my understanding is that venture capital also differs from other asset classes in that the best performers are much (and consistently) better than the rest. So, if asset classes have normal (or close to) distribution, venture capital returns certainly don't, so I'm not sure about the value of comparison. What were the returns for the top decile?

    Secondly, If legal structure of VC partnerships doesn't work, what about smart angel investors? They are not constrained by these structures, are they actually doing better or we just assume since we don't really know?

    Gabriela
  • Great comments, thanks Gabriela.

    Indeed, early stage investment returns are not normally distributed, but done well - ie systematically - the returns are not random but have distributions that do emerge over time. Like in many other walks of life, data is finally starting to emerge on this and I think with the advent of organisations like seedcamp, Y Combinator, TechStars and many others, you will start to see more and more data become available. I'd have to break out my 20 year old stats textbooks to be able to opine on whether these returns fit a known distribution, but I think it is fairly obvious to most observers that it is some kind of power law. The main point I was trying to emphasize is that returns in this asset class are structurally uncorrelated to most other financial assets, for basically two reasons:

    * growth/success of each individual portfolio company is (almost) entirely uncorrelated to the overall economy; starting from zero, you either grow or you don't. To be fair, if the economy is roaring or crashing, you might grow a bit faster or slower but this will be at best a distant second-order effect. Indeed in some cases there may even be a negative correlation as customers may be more likely to adopt a new, better, disruptive service or product when times are tough; when everything is booming, this can often be a challenge for a new company as customers may acknowledge their better mousetrap but can't really be bothered to switch from their current suppliers or way of doing business.
    * investments in these sorts of assets are illiquid with long (6-12 year) horizons; if you try to trade these assets on a shorter timescale, you will increase the correlation with other financial assets, ie if you need liquidity you will almost certainly erase (and perhaps more) any structural advantage inherent in the underlying strategy. Clearly you need to get paid for this illiquidity - and have the Buffet-esque nerves to stick to appropriate investment horizons - if you want to harvest these returns. I would posit that it is often here where investors in early stage companies fail.

    As for the structure of VC partnerships, I wouldn't say they don't work full stop, but that they are in fact ill-adapted to early stage investing. There capital and incentive structures and time horizons are misaligned. I would suggest that the typical VC fund structure is well adapted to investing in more mature growth companies - Series B, C, D in the industry jargon is probably their sweet spot. More early stage "private equity" so to speak. With respect to 'smart angels' indeed I think they are doing better and believe a significant part of this is due to the fact that they can align their personal 'structure' to the opportunity better. Hard data is still spotty, but a great study in the UK from last year estimated that angel returns over the last decade (in the UK) were on the order of 22% per annum (vs basically nought for VCs.)

    In conclusion, one point I really want to emphasise is that in any ecosystem a diversity of business models is typically more robust and appropriate than homogeneity, and that in terms of investing in private companies there is no reason this "truth" shouldn't hold as well. But human nature being what it is, industries tend to herd - eventually to excess - around one successful model and eventually "forget" why and in what context that model enjoyed it's original success. On a positive note however, out of this excess and diminishing returns, new opportunities arise and indeed I think that is exactly what we are starting to see happen, with the apocryphal "rise of the super-angels" for example being a manifestation of this.
blog comments powered by Disqus