Archive for the 'winner-take-all markets' Category

The Social Cost of Finance

Noah Smith has a great post that bears on the topic that I have been discussing of late (here and here): whether the growth of the US financial sector over the past three decades had anything to do with the decline in the real rate of interest that seems to have occurred over the same period. I have been suggesting that there may be reason to believe that the growth in the financial sector (from about 5% of GDP in 1980 to 8% in 2007) has reduced the productivity of the rest of the economy, because a not insubstantial part of the earnings of the financial sector has been extracted from relatively unsophisticated, informationally disadvantaged, traders and customers. Much of what financial firms do is aimed at obtaining an information advantage from which profit can be extracted, just as athletes devote resources to gaining a competitive advantage. The resources devoted to gaining informational advantage are mostly wasted, being used to transfer, not create, wealth. This seems to be true as a matter of theory; what is less clear is whether enough resources have been wasted to cause a non-negligible deterioration in economic performance.

Noah underscores the paucity of our knowledge by referring to two papers, one by Robin Greenwood and David Scharfstein (recently published in the Journal of Economic Perspectives) and the other, a response by John Cochrane posted on his blog (see here for the PDF). The Greewood and Scharfstein paper provides theoretical arguments and evidence that tend to support the proposition that the US financial sector is too large. Here is how they sum up their findings.

First, a large part of the growth of finance is in asset management, which has brought many benefits including, most notably, increased diversification and household participation in the stock market. This has likely lowered required rates of return on risky securities, increased valuations, and lowered the cost of capital to corporations. The biggest beneficiaries were likely young firms, which stand to gain the most when discount rates fall. On the other hand, the enormous growth of asset management after 1997 was driven by high fee alternative investments, with little direct evidence of much social benefit, and potentially large distortions in the allocation of talent. On net, society is likely better off because of active asset management but, on the margin, society would be better off if the cost of asset management could be reduced.

Second, changes in the process of credit delivery facilitated the expansion of household credit, mainly in residential mortgage credit. This led to higher fee income to the financial sector. While there may be benefits of expanding access to mortgage credit and lowering its cost, we point out that the U.S. tax code already biases households to overinvest in residential real estate. Moreover, the shadow banking system that facilitated this expansion made the financial system more fragile.

In his response, Cochrane offers a number of reasons why Greenwood and Scharfstein are understating the benefits generated by active asset management. Here is a passage from Cochrane’s paper (quoted also by Noah) that I would like to focus on.

I conclude that information trading of this sort sits at the conflict of two externalities / public goods. On the one hand, as French points out, “price impact” means that traders are not able to appropriate the full value of the information they bring, so there can be too few resources devoted to information production (and digestion, which strikes me as far more important). On the other hand, as Greenwood and Scharfstein point out, information is a non-rival good, and its exploitation in financial markets is a tournament (first to use it gets all the benefit) so the theorem that profits you make equal the social benefit of its production is false. It is indeed a waste of resources to bring information to the market a few minutes early, when that information will be revealed for free a few minutes later. Whether we have “too much” trading, too many resources devoted to finding information that somebody already has in will be revealed in a few minutes, or “too little” trading, markets where prices go for long times not reflecting important information, as many argued during the financial crisis, seems like a topic which neither theory nor empirical work has answered with any sort of clarity.

Cochrane’s characterization of information trading as a public good is not wrong, inasmuch as we all benefit from the existence of markets for goods and assets, even those of us that don’t participate routinely (or ever) in those markets, first because the existence of those markets provides us with opportunities to trade that may, at some unknown future time, become very valuable to us, and second, because the existence of markets contributes to the efficient utilization of resources, thereby increasing the total value of output. Because the existence of markets is a kind of public good, it may be true that even more market trading than now occurs would be socially beneficial. Suppose that every trade involves a transaction cost of 5 cents, and that the transactions cost prevents at least one trade from taking place, because the expected gain to the traders from that trade would only be 4 cents. But since that unconsummated trade would also confer a benefit on third parties, by improving the allocation of resources ever so slightly, causing total output to rise by, say, 3 cents, it would be worth it to the rest of us to subsidize parties to that unconsummated trade by rebating some part of the transactions cost associated with that trade.

But here’s my problem with Cochrane’s argument. Let us imagine that there is some unique social optimum, or at least a defined set of Pareto-optimal allocations, which we are trying to attain, or to come as close as possible to. The existence of functioning markets certainly helps us come closer to the set of Pareto optimal allocations than if markets did not exist. Cochrane is suggesting that, by devoting more resources to the production of information (which in a basically free-market, private-property economy involves the creation private informational advantages) we get more trading, and with more trading we come closer to the set of Pareto-optimal allocations than with less trading. However, it seems plausible that the production of additional information and the increase in trading activity is subject to diminishing returns in the sense that eventually obtaining additional information and engaging in additional trades reduces the distance between the actual allocation and the set of Pareto-optimal allocations by successively smaller amounts. Otherwise, we would in fact reach Pareto optimality. So, as we devote more and more resources to producing information and to trading, the amount of public-good co-generation must diminish. But this means that the negative externality associated with using increasing amounts of resources to produce private informational advantages must at some point — and probably fairly quickly — overwhelm the public-good co-generated by increased trading.

So although Cochrane has a theoretical point that, without more evidence than we have now, we can’t necessarily be sure that the increase in resources devoted to finance has been associated with a net social loss, I am still inclined to suspect doubt strongly that, at the margin, there are net positive social benefits from adding resources to finance. In this regard, the paper (cited by Greenwood and Scharfstein) “The Allocation of Talent: Implications for Growth” by Kevin Murphy, Andrei Shleifer and Robert Vishny.

Falling Real Interest Rates, Winner-Take-All Markets, and Lance Armstrong

In my previous post, I suggested that real interest rates are largely determined by expectations, entrepreneurial expectations of profit and household expectations of future income. Increased entrepreneurial optimism implies that entrepreneurs are revising upwards the anticipated net cash flows from the current stock of capital assets, in other words an increasing demand for capital assets. Because the stock of capital assets doesn’t change much in the short run, an increased demand for those assets tends, in the short run, to raise real interest rates as people switch from fixed income assets (bonds) into the real assets associated with increased expected net cash flows. Increased optimism by households about their future income prospects implies that their demand for long-lived assets, real or financial, tends to decline as household devote an increased share of current income to present consumption and less to saving for future consumption, because an increase in future income reduces the amount of current savings needed to achieve a given level of future consumption. The more optimistic I am about my future income, the less I will save in the present. If I win the lottery, I will start spending even before I collect my winnings. The reduced household demand for long-lived assets with which to provide for future consumption reduces the value of such assets, implying, for given expectations of their future yields, an increased real interest rate.

This is the appropriate neoclassical (Fisherian) framework within which to think about the determination of real interest rates. The Fisherian theory may not be right, but I don’t think that we have another theory of comparable analytical power and elegance. Other theories are just ad hoc, and lack the aesthetic appeal of the Fisherian theory. Alas, the world is a messy place, and we have no guarantee that the elegant theory will always win out. Truth and beauty need not the same. (Sigh!)

Commenting on my previous post, Joshua Wojnilower characterized my explanation as “a combination of a Keynesian-demand side story in the first paragraph and an Austrian/Lachmann subjective expectations view in the second section.” I agree that Keynes emphasized the importance of changes in the state of entrepreneurial expectations in causing shifts in the marginal efficiency of capital, and that Austrian theory is notable for its single-minded emphasis on the subjectivity of expectations. But these ideas are encompassed by the Fisherian neoclassical paradigm, entrepreneurial expectations about profits determining the relevant slope of the production possibility curve embodying opportunities for the current and future production of consumption goods on the one hand, and household expectations about future income determining the slope of household indifference curves reflecting their willingness to exchange current for future consumption. So it’s all in Fisher.

Thus, as I observed, falling real interest rates could be explained, under the Fisherian theory, by deteriorating entrepreneurial expectations, or by worsening household expectations about future income (employment). In my previous post, I suggested that, at least since the 2007-09 downturn, entrepreneurial profit expectations have been declining along with the income (employment) expectations of households. However, I am reluctant to suggest that this trend of expectational pessimism started before the 2007-09 downturn. One commenter, Diego Espinosa, offered some good reasons to think that since 2009 entrepreneurial expectations have been improving, so that falling real interest rates must be attributed to monetary policy. Although I find it implausible that entrepreneurial expectations have recovered (at least fully) since the 2007-09 downturn, I take Diego’s points seriously, and I am going to try to think through his arguments carefully, and perhaps respond further in a future post.

I also suggested in my previous post that there might be other reasons why real interest rates have been falling, which brings me to the point of this post. By way of disclaimer, I would say that what follows is purely speculative, and I raise it only because the idea seems interesting and worth thinking about, not because I am convinced that it is empirically significant in causing real interest rates to decline over the past two or three decades.

Almost ten months ago, I discussed the basic idea in a post in which I speculated about why there is no evidence of a strong correlation between reductions in marginal income tax rates and economic growth, notwithstanding the seemingly powerful theoretical argument for such a correlation. Relying on Jack Hirshleifer’s important distinction between the social and private value of information, I argued that insofar as reduced marginal tax rates contributed to an expansion of the financial sector of the economy, reduced marginal tax rates may have retarded, rather than spurred, growth.  The problem with the financial sector is that the resources employed in that sector, especially resources devoted to trading, are socially wasted, the profits accruing to trading reflecting not net additions to output, but losses incurred by other traders. In their quest for such gains, trading establishments incur huge expenses with a view to obtaining information advantages by which profits can be extracted as a result of trading with the informationally disadvantaged.

But financial trading is not the only socially wasteful activity that attracted vast amounts of resources from other (socially productive) activities, i.e., making and delivering real goods and services valued by consumers. There’s a whole set of markets that fall under the heading of winner-take-all markets. There are some who attribute increasing income inequality to the recent proliferation of winner-take-all markets. What distinguishes these markets is that, as the name implies, rewards in these markets are very much skewed to the most successful participants. Participants compete for a reward, and rewards are distributed very unevenly, small differences in performance implying very large differences in reward. Because the payoff at the margin to an incremental improvement in performance is so large, the incentives to devote resources to improve performance are inefficiently exaggerated. Because of the gap between the large private return and the near-zero social return from improved performance, far too much effort and resources is wasted on achieving minor gains in performance. Lance Armstrong is but one of the unpleasant outcomes of a winner-take-all market.

It is also worth noting that competition in winner-take-all markets is far from benign. Sports leagues, which are classic examples of winner-take-all markets, operate on the premise that competition must be controlled, not just to prevent match-ups from being too lopsided, but to keep unrestricted competition from driving up costs to uneconomic levels. At one time, major league baseball had a reserve clause. The reserve clause exists no longer, but salary caps and other methods of controlling competition were needed to replace it. The main, albeit covert, function of the NCAA is to suppress competition for college athletes that would render college football and college basketball unprofitable if it were uncontrolled, with player salaries determined by supply and demand.

So if the share of economic activity taking place in winner-take-all markets has increased, the waste of resources associated with such markets has likely been increasing as well. Because of the distortion in the pricing of resources employed in winner-take-all markets, those resources typically receiving more than their net social product, employers in non-winner-take-all markets must pay an inefficient premium to employ those overpaid resources. These considerations suggest that the return on investment in non-winner-take-all markets may also be depressed because of such pricing distortions. But I am not sure that this static distortion has a straightforward implication about the trend of real interest rates over time.

A more straightforward connection between falling real interest rates and the increase in share of resources employed in winner-take-all markets might be that winner-take-all markets (e.g., most of the financial sector) are somehow diverting those most likely to innovate and generate new productive ideas into socially wasteful activities. That hypothesis certainly seems to accord with the oft-heard observation that, until recently at any rate, a disproportionate share of the best and brightest graduates of elite institutions of higher learning have been finding employment on Wall Street and in hedge funds. If so, the rate of technological advance in the productive sector of the economy would have been less rapid than the rate of advance in the unproductive sector of the economy. Somehow that doesn’t seem like a recipe for increasing the rate of economic growth and might even account for declining real interest rates. Something to think about as you watch the Lance Armstrong interview tomorrow night.


About Me

David Glasner
Washington, DC

I am an economist at the Federal Trade Commission. Nothing that you read on this blog necessarily reflects the views of the FTC or the individual commissioners. Although I work at the FTC as an antitrust economist, most of my research and writing has been on monetary economics and policy and the history of monetary theory. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 245 other followers


Follow

Get every new post delivered to your Inbox.

Join 245 other followers