By Jessica E. Lessin

The “B” word returned this week. The stock market swooned. Bitcoin tanked. And the world was abuzz with conversations about whether bubbles were bursting—or not. 

Against this sudsy backdrop, I reached out to Bill Janeway, who has been studying bubbles as an economist and growth equity investor. Joining Warburg Pincus in the late 1980s, he backed enterprise giants like BEA Systems and Veritas and watched the 1998 to 2000 tech bubble up close. 

The stock market and bitcoin sell-offs are sparking a new debate about bubbles. Economist and investor Bill Janeway sees a bubble in both private tech and the crypto markets but says neither poses a systemic threat.

I asked Janeway for the history of bubbles and what it might teach us about the private tech and crypto markets. His answers, including excerpts from his lectures and book “Doing Capitalism in the Innovation Economy,” are below. He starts with some interesting thoughts about the difference between bad and benign bubbles and what to look out for with each. He then applies that framework to the present. It’s meaty stuff, so I have added some summary points in bold. 

Bubbles vary based on the objection of the speculation and whether banks or the equity markets are involved. 

History tells us to distinguish between bubbles along two different dimensions. One dimension is defined by the object of speculation. Only occasionally have speculators focused on fundamental technology, instead of such assets as gold mines or houses, neither of which contributes to system-wide increases in productivity.

The second dimension concerns the locus of speculative activity, distinguishing between bubbles that remain confined to the capital markets and those that transcend the capital markets to suck in the institutions that accept deposits and provide the credit that fuels the ordinary workings of the market economy. 



A crucial distinguishing factor is the degree of leverage, the relative magnitude of debt mobilized to finance the unsustainable rise in asset prices. In the trading markets for equities and bonds, it is limited: on the order of 1:1 at most. In the banking system as of 2007, it ranged upward from reported levels of 30:1 and beyond—in reality, far beyond: marginal declines in the value of assets rendered the institutions at the core of the financial economy insolvent. And these reported capital ratios substantially overstated the robustness of the banks’ balance sheets.

The dot-com bubble and the more recent credit bubble represent the two different types. 

The contrast is instructive between the respective consequences, positive and negative, of the dot-com/telecom bubble of 1998 to 2000 and the credit bubble of 2005 to 2007. When the $6 trillion of nominal financial wealth created in the former and concentrated in equity securities and junk bonds was liquidated, the economic consequences were within the bounds of post-war experience, leaving the technological foundations and business models in place for the newest new economy. 

The great credit bubble will be remembered precisely for its destructive economic consequences and not for any physical legacy, least of all the abandoned tract houses scattered along the coastal regions of the United States and from Ireland and Spain to the emerging lands of Central and Eastern Europe. The dichotomy echoes that between the limited economic consequences of the stock market crash of 1929 and the overwhelming impact of the international banking crises of 1931 to 1933.

Productive bubbles stimulate necessary investment in new industries. 

In theoretical terms, productive bubbles solve a coordination problem in time. New ventures that require substantial follow-on investment, such as technologically innovative infrastructure, are unlikely to gain initial funding except in an environment where the first round investors can be reasonably confident that the capital required will be available. 

Moreover, new ventures founded to explore the new economic space opened up by the new infrastructure are bound to be financially constrained, dependent upon external funding until they can reach critical mass and the corporate bliss of positive cash flow from operations. 

Financial bubbles focused on the potential of new technology relieve the financial constraint and enable the surge of network building and experimentation necessary for the construction of a “new” economy: whether based on railroads, electricity, highways or the internet.

The private tech markets and crypto markets are behaving like bubbles. 

On the one hand, the massive cash deficits of the “unicorns” of the digital revolution, offering the dream of yet another Google or Facebook, have been funded at valuations that represent premiums to their solidly profitable, public models. The longstanding reality that illiquid investments should trade at a discount has been reversed.

On the other hand, cryptocurrencies devoid of any fundamental link to cash flow, present or possible, have scaled the utmost heights of nominal valuation. In both cases, the telltale signature of a bubble is visible: Instead of rising prices choking off demand, rising prices have been accompanied by increased volume. The incipient return of positive real returns in the credit markets was always likely to dampen speculative fervor, and so it seems to be doing.

The private tech and crypto ‘bubbles’ don’t pose systemic threats. 

No doubt, some set of unicorns will find their way to positive cash flow, delivering valued services discovered as such thanks to the running room provided by speculative excess. Is there a sustainable role for crypto-currencies? Certainly the fantasy of digital “money” displacing government-backed “fiat” is beyond remote. 

The very volatility of the asset’s price renders it useless as a unit of account or medium of exchange; as a store of value, digital assets are proving to be marginal playthings. Moreover, the current state of the underlying blockchain technology is of limited use given its pathetic throughput and energy intensity versus conventional payment processes. 

Ironically, any productive application of the distributed infrastructure is most likely to emerge from the collaboration of large, trusted, visible, regulated partners—big banks and their clients—not from the shadows on the margins of cyber-land. Given the relative lack of leverage in both cases, neither bubble offers any significant, systemic threat.

published on February 9, 2018 on The information