Dispatches From The Dark Side

The Silicon Valley Diaspora

I wrote the following piece after returning from a trip to India in 2004.  This is as urgent an issue for Silicon Valley and the competitiveness of the US Economy today as it was then.  While I wrote this about India, we need to be just as concerned about China and Eastern Europe.

February 23, 2004 

Silicon Valley is suffering a brain drain that is just going to keep getting worse.  Historically the valley has been the beneficiary of brain drains from around the world.  The diasporas of China, India, Russia, etc, all came to sunny California to start their companies and make their fortunes.  Companies like Juniper, Q-Logic, Exodus, and Silicon Labs are among the hundreds of companies that have generated billions in sales and wealth started by the best and brightest from around the world.  Now some of those great minds are packing up and heading home.

I spent the first week of February in India with a TiE delegation of other venture capitalists and TiE Charter members.  Everything you’ve heard about how exciting and dynamic things are in India today is true – if anything, it’s understated.  Politicians and labor unions who are expressing concern about US call center and SI jobs going to India are focusing their energies on the wrong problem.  The problem is not the inevitable transfer of certain business functions to where they can most efficiently be performed, but the migration of the entrepreneurs and experienced workers who create the companies (and the wealth) to their ancestral homelands.

We’ve all heard the stories about the quality of the Indian workforce.  They’re smart, they’re well educated, and yes, compared to engineers in Silicon Valley, they’ll work for peanuts.  But there are two critical pieces of human capital that India (and countries like it) lack that Silicon Valley has in spades.  The first is years of relevant experience – each new generation of technology building on the successes and failures of the one that came before it. The second is the high-tech entrepreneurs – the people who have worked in an industry for long enough to know how things are done and to know where the problems are that are worth solving.  But there’s nothing genetic that leads these shores to produce more entrepreneurs than anywhere else.

For the last 30+ years, the best and brightest from around the world came here.  They first came here to get educated.  That’s why the graduate programs at our top engineering schools are filled to overflowing with the best and brightest from India, China, Russia, and everywhere else around the world.  The world’s Diaspora came to the US to get educated.  And then once educated, they realized that the companies that would put all that good education to best use were companies right here in the good ole U-S of A.  If they happened to be in an area where companies are started as fast and furious as they are in Silicon Valley, they may have gotten the idea and tried to start one themselves.  They built up that critical experience and some even became one of our most treasured natural resources: the entrepreneur who innovates, creates jobs, and creates wealth.

Now some of those people are heading home.  So now those bright engineers who graduate from IIT with arguably the best engineering education in the world won’t be forced to reinvent the wheel.   They’ll have mentors who learned their trade through years of hard work in the valley and elsewhere.  And some of those that head back will be the entrepreneurs with the ideas and the initiative to start high-tech companies that are ready to compete on a global scale.

We now have our own Silicon Valley Diaspora to moan about - the best and the brightest leaving these shores for bigger and better opportunities elsewhere. So watch out.  While the first wave of Indian high-tech startups was dominated by call-center and IT outsourcing firms, the next wave is likely to be dominated by software and hardware companies that build their entire companies – not just their call-centers and back-office – somewhere other than Silicon Valley.

The JOBS Act Is Good, But SarbOx Shouldn't Get All The Blame

The JOBS Act is a good thing for Silicon Valley and the whole startup ecosystem.  Josh Kopelman is correct when he points out that IPO's over the last 10 years have happened later in companies' development than was previously the case and this has taken from public market investors the opportunity for tremendous value creation.  Dan Primack also makes a valid point that VC's have three options for exiting their portfolios: sell them, take them public, or shut them down.  But making SarbOx the sole scapegoat for why IPO activity has been anemic since 2000 is just as wrong as claiming it's the savior that has kept the public markets from suffering losses from failed companies that manage to go public.  Like @pkedrosky, I still believe Groupon will end up with gnashing of teeth and tears for investors of all stripes.

It's an oversimplification to blame the IPO market of the 2000's on SarbOx.  To understand the market during this period, you have to first remember what happened before it.  Anyone who was a part of the startup ecosystem during the bubble was, at least up until now, permanently formed by their experience.  It's like my grandfather, a child of the great depression, who 60 years later was still untrusting of banks and more frugal than anyone 10 years his junior.  The memories of public market investors aren't quite as long, but the hangover from the crash certainly lasted well into the 2000's.  When you think of how high the bar was for tech companies to go public post the bubble, just remember how big Google was when they finally went public – they were almost $1B in revenue in the fiscal year before their IPO.  A size that was completely unheard of in the venture-backed IPO world prior to Google.  In fact, the year that SarbOx passed, Google's revenue was $348m – a number that was well beyond what was needed to go public for any venture backed tech company up until that time.

At the time of Google's IPO, many were hopeful that it would usher in a new era of venture-backed IPO's.  Sadly, it did not.  And while the hangover was partially to blame, and SarbOx played a role, there was more to it than that.  Three other major transformations of the landscape, both before and after the bubble, have helped to make early venture-backed-IPO's more than a little challenging.

The first of those challenges was the "death" of the Four Horsemen.  During the 80's and early 90's, the bulk of venture-backed IPO's were done by four west-coast boutique investment banks – Robertson Stephens, Alex Brown, Hambrecht & Quist, and Montgomery Securities.  These banks made a great business out of doing smaller IPO's and the then denizens of Sandhill Road didn't see using them, versus a bulge-bracket firm, as a sign of weakness.  The late 90's changed all that with the larger IPO's that were able to get done during the bubble.  Once the fees got bigger, the bulge-bracket firms swooped in.  As they did, it became a badge of honor to have a bulge-bracket firm lead your IPO squeezing out the boutiques.  Over time, the Four Horsemen were all bought and no boutique-banks have managed to fully take their place – at least not to the extent that the Denizens of Sandhill road will allow them to lead their companies' IPOs.

The next two challenges were more structural changes to the market.  The first of those was the Global Settlement between Elliot Spitzer and the 10 largest investment banks, a topic Henry Blodget is all-too familiar with.  In an attempt to remove the inherent conflicts of interest that existed between investment banking and research at these firms, they managed to throw the baby out with the bath-water. What small IPO's need is marketing, and research is marketing.  If you try to take a company out and no one is talking about it, no one will trade it.  That tends to tilt the field towards companies going public only after they have become household names which usually implies much larger companies.

 The last nail in the coffin (not necessarily in chronological order) of smaller venture-backed IPOs was decimalization.  In late 2000, the NYSE started to move from trading in 8th's to trading based on decimals.  The biggest impact of this was on the spread between the bid and the ask.  If you're trading in 8ths, the spread scale can be 12.5 cents at a minimum and go up from there and when you trade in decimals, the spread between the bid and the ask can be as little as a penny.  Market makers, the guys who keep a stock trading, make their money on the spread between the bid and the ask.  And when that spread is squeezed, they make less money.  This is all good and well for a company like Apple who trades upwards of 20m shares a day, but for a small-cap stock that trades <100k shares a day, what they really need are market makers.  And if you can't make money on the spread, who wants to make a market in the stock.

 There's a lot there, and I'm sure there are other subtle changes in the market and landscape that I've missed. But on the bright side, we're finally getting over the hangover and the JOBS Act may have come along right at the perfect time.  There are more and more players every day who got into the game after the bubble.  What we need now is for a few of them to take their companies public early and be successful for public market investors.  Once that happens, all these challenges will wash away and we'll have a healthy venture-backed IPO market again.