Doing Capitalism in the Innovation Economy by William Janeway

Bill Janeway has a lot to say in this book, and he does so with dense and clear prose. In three points, this book makes clear:

  • The innovation economy is a three player game among the state, the market economy, and financial institutions. The game is played with irreducible uncertainty, and playing by the rules of homo-economicus (some neoclassical rational agent) is not a successful strategy.
  • Speculation and state funding are essential to pushing out the knowledge frontier - waste and tumult should be expected and are needed to fund ideas that may not be cash positive in the near future.
  • Do not expect this process to be orderly; the most valuable innovations accompany a large amount “Schumpeterian waste,” the money and time plowed into speculation and dead-end investments. Schumpeter’s theory of creative destruction in which waste and excess play a prime role in disrupting entrenched players is key.

While he covers a wide variety of topics in detail, Janeway spends a great deal of time showing how these themes are viewed through the lens of investment strategy, bubbles, and the State’s role in the economy.

Investing

The architecture of investing has been uprooted over the last few decades - regulatory changes, decreases in fixed brokerage commissions, institutionalization of savings. Each change has further pushed firms to act as principals instead of agents, and the shift in incentives and compensation generation had a direct impact on the financial system. Securitization allowed firms to originate and transfer assets for huge profit while increasing dependence on both funding liquidity and market liquidity. Any new business entering this financial arena should have clear eyes about where incentives align and diverge.

For business management, “Cash and Control” are the essence of liberation from the constraints of competitive markets and economic efficiency. Cash buys time to figure out what is going on; Control permits the player to use that time to shift the parameters of the problem. The biggest winners in the innovation economy choose no financial risk at all. While winners take all shapes and sizes, there is a consistent formula that produces results: demonstrate competence to gain trust, leverage the trust for capital and control, redound on demonstrating competence in new role, repeat. Janeway points to JP Morgan as an example of this cycle, but countless examples can be found.

For an investor looking at a business, there is no notion of “right” valuation. Discounted cash flow, technical market metrics, and potential sale price all are dependent on judgment; predicting the future is embedded in the number produced. When there is a widespread consensus, we see steep discounts or premia, but uncertainty has not disappeared. Financial markets exist to intermediate the uncertainty, expressed through differences in opinions and valuations. Aligned incentives and trust leads to access to capital, the gate that gives investors the ability to pounce when prices are in their favor. The investor is not pitching a specific investment to the capital holders but pitching their own judgment to access capital freely and deploy it strategically.

How do businesses in the innovation economy differ from others? Janeway offers a number of useful mental models used to segment companies and businesses:

  • Projects vs. Products: Projects are owned by the customer, have an expiration date, and are custom built for each client. A product requires an ongoing relationship between customer and company, customizable but not custom for each client.
  • State pipeline: There is a network connecting state funded science, research commodified into product, and private demand for this product. Biotech has been a focus of VC following this model - state funded pipeline of research ready to be deployed into treatments, demand is funded by insurers and is inelastic, TAMs are known from the outset, and risks beyond scientific are estimable.
  • Innovator’s Dilemma: an original innovator now with surplus resources is crippled by its own inability to cannibalize its margins with innovations that will eventually take them anyway. Janeway focuses on the early computer industry, where disruptions were plentiful and big players like IBM stumbled to adjust to open source protocols. This point seems tied to commoditizing your complement; finding the points of leverage (or Power in Helmer’s terminology) in complementary components to your product and turning them into interchangeable parts.
  • UI vs Tech: The best technology does not always win, and often fails to win. Customers do not see or care about the best technology, but want the best product. Often the founders with the best technology dismiss the importance of user-interfaces, marketing, and sales.
  • The Paradox of Prudence: In tough times, companies shrink their balance sheets and households tilt portfolios away from real investment towards safe assets. This is micro-prudent but macro-imprudent as it makes the economy more risky as liquidity retreats. Attempts to reduce individual risks raises endogenous risks. Like Keynes’ Paradox of Thrift - individual increases in savings can reduce aggregate savings by reducing economic activity.
Fun Detail
  • Venture capitalist Fred Adler adds some levity to the narrative and has some choice quotes.
    • “There is no such thing as a fixed cost.” Every cost is variable in the long run, and the only question is simply how much money and time it takes to turn a fixed into a variable cost.
    • The definition of corporate happiness: “Positive cash flows from operations.” This fits neatly with the broader idea of Cash and Control; without positive cash flow, you do not control your own destiny.

Bubbles

Janeway’s dual role of investor and economist shines in this chapter, as he connects economic models to his lived experience. Key to this chapter is his use of the Rational Expectations Hypothesis (REH), the idea that financial markets behave as if they were complete and efficient through the mind of a rational representative agent, one who knows all relevant information. Add to this model the Efficient Markets Hypothesis, and the market will efficiently allocate capital to the best ventures. With these two ideas, we can construct the idea of the rational bubble: a continuous rise in the asset’s price leads investors to be content to hold at the new valuation, as the risk for the bubble bursting is offset by potential rate of price increase. Bubbles can have rational behavior at an individual level while still exhibiting a full coordination failure of the market. On the other hand momentum driven behavior, increased buying as prices rise and increased selling as prices fall, goes against what we might expect under REH.

Market bubbles are often accompanied by moralizing or the assumption that an investor should position defensively. Janeway has no qualms demonstrating how an investor can take advantage of the upside as well, and shows how Warburg Pincus benefitted from price bubbles in their holdings. By recognizing the bubble, the firm was able to divest and distribute towards the peak of valuations. However, one must not cloud personal judgment by substituting the market’s enthusiasm for one’s own projections.

Bubbles certainly have their downsides. Financial speculation can have a direct negative role on the innovation economy, discrediting innovative technology and retarding its progress as it struggles to find funding after the crash. As Hyman Minsky noted, the Fed steps in to prevent catastrophe and maintain markets, validating the past use of a risky instrument and serving as a guarantee on future speculation, inevitably requiring another intervention down the road. Minsky’s economic model, long overlooked before gaining notoriety in the aftermath of the GFC, posits that the economy naturally swings between stable and unstable financing regimes. Over periods of prolonged prosperity, the economic momentum alters financial relations towards levels of indebtedness that threaten the real economy. The credit system rides waves through stages of confidence and risk taking from hedge finance to speculative finance to Ponzi finance.

But just as often, bubbles are essential to the pushing the boundaries of the innovation economy. Speculative investments that fail in the short-term provide a path towards success in the long-term. Productive bubbles differ from the purely financial ones; the bubble in US railways in the late 19th century funded massive infrastructure investment with tracks laid across the country. The crash may wipe out equity holders, but the real assets remain to be utilized by the next generation of companies.

What separates good bubbles from bad? There is a distinction between the object of speculation and the locus of speculative activity; the object defines the assets that inflate in value while the locus defines how widespread the exuberance extends. In other words, is the speculation confined to the capital markets or does it suck in depositors, credit extension, and systemically important institutions. The degree of leverage can determine whether a bubble is productive or destructive, since following Minsky’s model large defaults and restructuring make the locus of losses likely to spread beyond the capital markets. We can view 2000 as a productivity enhancing bubble confined to the equity markets and 2008 as a credit driven bubble spread throughout the banking system with negative effects for the real economy.

The State and Macroeconomics

Here the economist Janeway is in full force, and many theories of state power and the role of capitalism are contrasted. The government plays a key roll in fundamental economic processes of growth, like promoting

  • Keynesian waste: deadweight loss represented by unemployed resources of human labor and physical capital. Keynes worries economic inactivity has long-term effects on productivity as skills atrophy and capital depreciates. It is preferable to employ workers in completely wasteful engagements, as maintaining zero productivity growth is better than a permanent loss of output potential.
  • General purpose technology: widely used, capable of ongoing improvement, enables innovation in application sectors. Basic scientific innovations build thick stacks of technologies and applications.
  • Spillovers: technology spillovers increase the productivity of other firms that operate in similar areas.

While the specific points outlined are interesting on their own, we may still ask what kind of state generates productive investment? Janeway sees several key roles for the state. First, strong patent policies, along with protective tariff policies for fledgling economies, provide the bedrock for private investment and enterprise. Second, the state is there to support fundamental research into new technology, though this point is not without political controversy. Janeway points to the left, which often sees new technologies as job destroying.

As a US example, world wars supplanted private investment in research with defense-led initiatives and changed corporate patent structure by requiring major corporations to share results of research with each other and new entrants. Here the state fronted the cost of knowledge generation and made sure society benefitted through diverse commercial applications of the science. In contrast, much of Europe followed a model of protection for “national champion” companies, with quite apparent differences in results.

Finally, there is good point raised about the limits on the value of economic models. The macro economy is not something that any of our models easily predict, summarized as the “database problem.” Econometric modes are reliant on historical data to predict future moves, yet if we are living outside of the support of the data (eg. behavioral relationships are non-stationary) then our predictions may have no value in the current environment. This is where simulation is key; try to model agent-based incentives which should be stationary and simulate responses given possible states of future reality. This idea is reminiscent of Thomas Schelling’s Micromotives and Macrobehavior, which shows how structural problems can be understood through individual incentives.

Second, not only can we fail at estimating the parameters in our model, but we miss the complex structural relationships that can lead us to false conclusions. Ricardian equivalence, a model in which state borrowing is economically equivalent to state taxation since all actors know borrowings must be returned out of future tax revenues, is a fine example of over-reliance on narrow minded rational-agent models. This idea suffers from the unlikely assumption that all potential resources are already employed, so state spending comes without a multiplier. Keynes instead posited that state spending promotes the use of unused assets, boosting economic activity and potential state revenues at the same tax rates.

Conclusions

Doing Capitalism is a nice complement to M. Mitchell Waldrop’s The Dream Machine, turning a vivid example of government backed science into a fully developed economic theory. While State money kicked off the internet revolution, the privitization of innovation described in the The Dream Machine has only grown. The US government shows less interest in funding scientific corporate initiatives, and political incentives aren’t aligned toward producing the type of waste Janeway prizes (think Solyndra). However, private development has also taken a turn in the machine learning age, with the largest research companies offering their algorithms via API for free.

While popular business models have evolved (positive cash flow from operations is not exactly in vogue), Janeway provides a valuable mental model for understanding our modern economy and how the next innovation jump might arise. The exuberance of the crypto bubble left many with nothing to show for their investments, but it also introduced new funding mechanisms for distributed teams, incentivized developers to learn the new language of smart contracts, and opened experimentation in trustless systems. It’s too soon to declare the craze a productive bubble, but at the very least the crash failed to spillover into the real economy. The State also shows some interest in returning to the funding of science with an RFQ for tech that requires true advances. As a buyer of private space services and with the spectre of green investment in the future, the federal government could once again act as a scaffold to push the innovation frontier out further.