Why the Silicon Valley Model Doesn’t Work in Japan

John Sasaki, Attorney at Law, talking before ACCJ members at the Tokyo American Club in Tokyo

John Sasaki, Attorney at Law, talking before ACCJ members at the Tokyo American Club in Tokyo

Foreign entrepreneurs and investors who want to follow the Silicon Valley venture capital model in Japan can do so, but it will cost a fortune, raise tax risk, and might not work as expected. John Sasaki, a Silicon Valley lawyer based in Tokyo for the past 15 years, explains why.

The Silicon Valley VC model evolved to make it easier for entrepreneurs lacking money to start innovative new businesses. Without the Valley model, founders would pay the same price as investors for company shares. Much wealthier investors then end up owning most of the company. Everyone loses, because founders owning little equity lack motivation to champion the business.

To solve that problem, the Valley created another class of potentially more valuable shares for investors called preferred stock. The higher priced preferred stock carries special rights, allowing investors to reach risk-reward equilibrium with the founders. Preferred stock was worth many multiples of common stock. The actual multiple was determined by negotiation, as startup companies are difficult to accurately value. Eventually a 10 to 1 multiple emerged as the rule of thumb. The US’s Internal Revenue Service accepted the 10 to 1 rule, preferring not to second guess the judgment of those involved.

According to Sasaki, a similar rule of thumb never arose in Japan. Here, tax authorities appear to be much more likely to question a firm’s share price. (America’s IRS has recently become more aggressive in requiring justification, but tax authorities in Japan remain comparatively much more aggressive.) Their attitude seems to be, “Prove it. If you can’t, we’re going to tell you what it is worth.” Another reason is that Japanese investors have never risked testing the model. “Nobody wants to be the test case, so nobody knows if the Silicon Valley model of stock valuation works,” says Sasaki.

Just as important, nobody knows if all of the rights attached to preference shares are legally enforceable. Here, all shareholders must be treated equally by law. While US holders of preferred stock are guaranteed a return on their investment before founders get paid anything, lawyers cannot guarantee that investors in Japan will get their money back first. “They might, but then they might not. It’s an enforceability issue,” explains Sasaki. “How many foreign VCs are going to be investing in Japanese startups without a guarantee?” he asks. Not many, he suggests.

The Silicon Valley model is based on founders and key staff committing to work in the business for at least four years. If they fail to fulfill their commitment, then the company takes back some of their stock. There is a one year cliff, meaning that founders who leave within the first year get nothing. Those who get past the first year keep 25% of their shares. After that, shares vest on a monthly basis. The company also has the right to buy back shares from those who leave within four years, based on the time spent at the company. For example, the firm can buy back half the shares of founders who leave after two years. Those who remain all four years get to keep all their shares.

Vesting is widely used as a founder incentive in the US. In Japan, however, it is less common. The reasons are largely cultural. In the past, founders personally guaranteed their businesses to get needed funding. Perhaps they have developed a sense of entitlement from putting so much at stake? For whatever the reason, founders in Japan normally get to keep all their shares, regardless of when they quit.

Japanese employees also prefer cash over stock options. Given the choice between taking a job for $100,000 at Sony versus $50,000 plus stock options at a hot startup, a top engineer usually prefers to work for Sony. “That’s not a legal issue. It’s cultural,” says Sasaki.

The Silicon Valley model further depends on the ability of investors to profitably exit from firms they invest in. Consider the VC which invests in ten startups. One might turn out to be a ‘star’ which, through an initial public offering, recoups the losses incurred by, say, five ‘dogs’. The VC profits by selling the remaining four ‘question mark’ investments. That is easier to achieve in the US than in Japan.

In the US, 80% of startups which exit do so by sale to a larger company. Firms like Facebook and Google aggressively buy innovative startups in the hope that one will become a hit. In Japan, however, only 20% of exiting startups are sold to bigger firms. That’s because large Japanese companies prefer to develop technologies in-house, having developed the staff and research facilities during the post-war years under the tradition of lifetime employment. With few buyers, VCs have difficulty selling their ‘question marks’. “If they can’t sell them, the Silicon Valley model doesn’t work,” notes Sasaki.

Tax risk also thwarts acquisitive foreign firms from buying Japanese startups in the way they are accustomed (by triangular merger). In the case of a domestic buyer, the shareholders of the target company can defer tax liability through a share exchange agreement (kabushiki koukan), a method of acquiring a company which cannot be used by foreign firms. In the case of a foreign buyer, the transaction most likely results in a tax liability for target company shareholders. Foreign firms can alternatively buy company shares direct from individual shareholders. However, this can be a burdensome procedure, especially if the target company has many shareholders.

Unlike in the US, tax and company laws in Japan do not incentivize entrepreneurship. Here, more bureaucratic and patronizing tax authorities hinder people from building great companies like Facebook, Amazon and Google. More importantly, Japan’s risk-averse culture discourages entrepreneurship.

Sasaki believes the Silicon Valley model might not work in Japan, even if the tax issues and legal uncertainties were removed. What works in one culture may not work in another— equity incentives are a case in point. Sasaki asks, “Should Japan really be trying to replicate the Silicon Valley model?” That might take a generation. Alternatively, he thinks it would be better to adopt a model which actually motivates the Japanese.

Recent articles:

Beacon Reports reveals Japan through the lens of thought leaders. Subscribe free!

2 comments to “Why the Silicon Valley Model Doesn’t Work in Japan”
  1. To be fair, the Silicon Valley model has not worked almost anywhere else.
    The issues above are real, thank you Mr. Sasaki for the expose, but the Silicon Valley model is very possibly a first mover advantage which is not replicable.
    A few of the possible reasons:
    1) There’s so much capital in Silicon Valley that talent as well as entrepreneurs are drawn from all over the world to start companies, grow and get funded there
    2) Local markets are different – not just in the cultural and regulatory regimes noted above, but in consumer behavior. The US being a huge market and Silicon Valley being a leader for the technorati all over the world makes for a potentially stifling ROW ecosystem.
    3) The preponderance of huge technology companies in Silicon Valley prejudices startups to begin there as acquisition has long since replaced M & A as exit strategy
    No doubt there are more.

  2. I totally agree that in order to build an entrepreneurship culture here in Japan on par with Silicon Valley’s, we’ll need to first work on developing one that motivates that Japanese. That should have been the gist of the article.

    HOWEVER, the rest of the article is like a typhoon-hit exposition coastline, a depressing scattering of craggy facts and broken logic:

    1. The article implies that in Japan, the system is setup so that seed capital-seeking founders automatically lose control (ownership) at the hands of their wealthy investors. There are no such mechanism making this the rule. It is up to the founders to decide how much ownership they are willing to relinquish in exchange for seed capital. Naturally, the founders predetermine their share of the startup so that their team represents a majority, reserving a minority for angels and investors. The problem here is not knowing the ropes, not preferred stocks …

    2. Preferred stocks existed a few years before California even became a state, in as early as 1836 in Maryland, and was not invented to alleviate the problem proposed here.

    3. The “Companies Law of Japan” (Yoshii, AM&T 2008), points out that the class of preferred stock was to be formally introduced in Japan in 2008, meaning non-enforceability should not be an issue for investor who’d want to get into a startup via preferred stocks.

    4. I don’t know what the statistics say regarding what SV investors prefer in exchange for seed money. Is it preferred stock? I would think that would be a minority, with the majority interested in huge capital gains, rather than steady dividend payouts. What’s the point about preferred stock in this article anyway?

    5. “Firms like FB and Google aggressively buy innovative startups in the hope that one will become a hit.” No. They buy them for their perceived value which should enhance the acquiring company’s mission, and therefore, increase the overall value for shareholders. Most of what Google bought was to incorporate into itself with a few, like Youtube, already hits when they were bought.

    6. Though Mr. Sasaki has excellent credentials, I was expecting a better analysis rather than what sounds like a very casual case study or personal observation.

    There were a lot of informative pieces of data, but the logical ties were quite frankly sloppy. Please redo this article. The topic is of deep, personal interest to me.

Leave a Reply

Your email address will not be published.