The Financial Times this week had a lengthy piece “When the tech boom met reality” detailing how truly galactic levels of venture capital funding have produced little innovation and are now leading to truly galactic levels of losses. It is now clear that the “all point in one direction and throw money at it” approach to innovation does not work. Since technological innovation is key to our future prosperity and indeed happiness, it is perhaps worth reviewing how better approaches in the past have produced better results. Structural changes are needed.
The collapse of this VC boom is not surprising – the same thing happened in 2000-01. Only the scale was larger this time; venture capital funding invested in startups in 2021 was $330 billion, compared with only $100 billion at the 2000 peak. Both the surge in funding and its results were predictable, and indeed predicted in this column. Unsophisticated institutional investors, such as pension funds and university endowments see the newly attractive returns on venture capital, flood into funding VC investors, and swamp the market, producing endless copycat venture capital deals that compete with each other and prevent any of them from reaching the scale needed for profitability.
There are a few differences. This time around, the investee companies mostly have substantial revenues, unlike Pets.com and the losers of 2000 (and Pets.com, dozy concept as it was, has produced several substantial descendants in the easy-money era of the last few years; such is the unoriginality of venture capital investors). The other major difference is that the investee companies tend to stay private for round after round of venture capital funding rather than going public. This is partly due to the onerous requirements of the 2002 Sarbanes-Oxley Act for small public companies and partly because, once you have been bullied into submission by VC funds, there is little more they can do to make your life even more miserable and so you might as well return to them for more funding, rather than seeking spurious independence in the public markets.
The experiences after 2000 and recently, together with the steady decline in small business formation since the 1970s, strongly suggest that the current venture capital model does not work as an incubator for innovation. The U.S. market for bank financing is far too concentrated, with the mergers since the 1980s having conglomerated the financial sector into a few behemoths in even fewer financial centers. On the other hand, the venture capitalists, while doubtless diverse in ethnicity, sex and gender are very un-diverse indeed in their thought processes, as well as being almost equally geographically concentrated.
There are two types of innovation that need to be financed. By far the most important is micro-innovation, where a small company or even a lone inventor toils, often for years, at a new process and eventually makes it marketable. There is also however a subset of innovations that require vast resources to develop and tend to be very long-term in nature. These are best nurtured through large corporations, but only if those corporations can be weaned off short-termism – the Bell Laboratories of the 1950s is the ideal, but that requires 1950s capital markets and incentive structures, to which we must seek to return.
For the great majority of innovations, small-scale in nature at least initially, the greatest of them do not make the most money – the market for their innovations is limited at first, the time to develop them is too great and the chances of the initial entrepreneur making a mistake are too high. Dan Bricklin’s spreadsheet, VisiCalc, the first of its kind, did not make him a billionaire even though it revolutionized business practices; subsequent entrepreneurs made all the money. Even in the Gilded Age, the greatest American inventor of all, Thomas Edison, guessed wrong on the form of electric current that would be used in transmission, and ended only modestly wealthy.
The current system, whereby dozens of greedy young people are given a lottery ticket to billions of dollars by the private equity industry makes very little sense; it favors the corner-cutting fast-buck artists over the truly innovative, it tends to weed out eccentrics (and therefore geniuses, who tend to be socially impaired) and it is far too short-termist in its approach to innovation.
At the other extreme, state-funded innovation is hopeless; it is politically determined, incredibly wasteful and rarely leads to anything useful. Louis XIV’s Machine de Marly is a case in point. It took 7 years to build and cost £300,000 1680s pounds, the most complex machine of the 17th century, yet its purpose was only to feed the fountains at Versailles, which it was never able to do satisfactorily. In more modern times, you can also consider the Manhattan Project and the Apollo Space Program, both of which technically successful projects were inordinately expensive and have yet to yield us a reasonable return, more than half a century later.
The archetypal innovator from the Industrial Revolution was not James Watt, who merely made some improvements (admittedly critical) to a steam engine that had already been invented, and then profited greatly by teaming with a capable industrialist Matthew Boulton and playing dubious games with his patents. Instead, it was Thomas Newcomen, who invented the first working steam engine with moving parts in 1712 after a decade of attempts in his modest Devonshire workshop (earlier attempts at steam pumps, like that of Thomas Savery, had no moving parts and were thus mere coffee pots)!
Unlike Watt, Newcomen did not die even modestly rich, nor did he exploit patents; instead he was exploited by that of Savery, for a quite different device, which had been extended to 1733, several years after Newcomen’s death. As a result, Newcomen had to pay Savery’s patent trustees spurious royalties on every steam engine sold. Nevertheless, his engines, which needed to be very large to be mechanically efficient, working as they did only with atmospheric pressure, were sufficiently successful that by the patent’s expiry in 1733, 110 had been sold (more than double the number of steam engines in France as late as 1815). Total sales of Newcomen-type engines were around 1,400 by 1800, with some sold as late as the 1790s to avoid the greedy Boulton’s exploitation of the Watt patents,
Apart from not dying rich, Newcomen was about as different as it is possible to be from Bill Gates or Mark Zuckerberg; as an early buyer, the coal mine tycoon Sir James Lowther, wrote:
“We have been very successful from first to last in the timing of things about the Fire Engine, which I should hardly have entered upon if I had not met such a very honest good man as Mr. Newcomen, who I believe would not wrong anybody to gain ever so much.”
Does that sound like a man who would get financing from the modern venture capital sector? (If he did, they would steal the company off him in months!) To finance innovation, we need a system that finances a Newcomen, an unworldly inventor with an idea of genius and the skill to realize it and that works with Newcomen over the many years of improvement and modification needed to make the idea successful.
Fifty years after Newcomen’s invention, that system came into being, with the country banks. By the Bank of England Act, 1708 no bank other than the Bank of England itself was allowed to have more than six partners. For half a century, this legislation stunted the growth of financial services outside London, then after 1750 a plethora of provincial country banks arose, set up by local merchants and tradesmen, mostly with less than £10,000 of capital. By 1780 there were over 200 such banks, whose notes were accepted by each other at par (there was a mechanism of London clearing houses to achieve this). These country banks provided discounting and receivables financing to local business, and occasionally overdraft facilities (secured by a note of credit for twice the amount of the overdraft) and issued small-denomination notes when coin was scarce. They also acted as networking agencies; they would not finance equity or long-term loans directly, but their partners and wealthier customers might do so.
As a result, any working man with a good idea could talk to any of several country banks within a day’s horse ride of his business; there was little exclusiveness as country bank partners were not socially grand like the later merchant bankers. Through this mechanism the Industrial Revolution was financed; innovation slowed only after the middle 19th century when the banking system, through misguided legislation, began to agglomerate.
Innovation is best financed by a network of financial institutions that are highly localized, small and above all, intellectually and socially diverse — the 18th century country banks were ideal for this purpose. We need to recreate such a system (the U.S. banking system before 1933, with banks doing investment banking and forbidden to bank interstate, got quite close to it). Add to this a few corporate behemoths with incentives set up for long-term growth and diversification, prohibited from repurchasing their shares and able to set up laboratories like Bell, and you would have a system of innovation encouragement that functioned very much better than the present mess.
It is by far the best way to get out of our current trap of negative productivity growth.
(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)