mercoledì 24 febbraio 2010

The doomsday cycle

This VoxEU article shares many similarities to my previous posts. It doesn't mean I've been copied (I think no one has read any post yet), but that it is very interesting.

Financial regulation and monetary policy - (3)

While safety nets are probably much more harmful than useful, as they create the illusion of stability and incentize instability-creating risk taking in the long run, the case for or against regulations is probably more complicated.

A distinction could be made between actions which add to risk without adding to efficiency, which could be regulated with no opportunity costs in terms of foregone output; and actions which add to risk but also add to efficiency. In the latter case, regulations will impact both bad entrepreneurial decisions induced by moral hazard and good ones which result in a better allocation of productive resources: distinguishing between the two is necessary to minimize the output loss due to the quest for higher stability, but involves a knowledge problem, as market information about risks and returns are distorted, by assumption, and regulators cannot use them as guidelines.

Prudential regulations avoiding the former type of market actions are good for the economy, whereas prudential regulations of the latter type will be difficult to evaluate, and probably harmful for growth and efficiency. Public choice problems could be added to the bill, but I would like to highlight another potential problem which is more seldom taken into account: the very same "market failure" which induces entrepreneurs to make bad decisions will also induce entrepreneurs to circumvent regulations hampering his ability to exploit these profit opportunities.

If we consider the probably most relevant case (relevance is a concept that cannot be derived a priori but it is always relevant in the assessment of real market phenomena), i.e., "market failures" induced by safety nets, which we may dub "policy market failures", we have some sort of a schizophrenic policy.

Let's call "Ben" the source of policy moral hazard and "Dolly" the regulator (in the US, there are scores of regulators, and Dolly being the cloned sheep, I guess the name is quite apt). Ben induces policy moral hazard and entrepreneurs start behaving as masochistic morons; Dolly sees that there is something wrong with the process and starts regulating what entrepreneurs can do in order to limit their stupid behavior. Ben, however, will keep on inducing masochism on the market process by socializing risks and credit costs, and so entrepreneurs will have to forego profits in order to obey Dolly: the most likely result is that they will try to circumvent, and sometimes maybe corrupt, Dolly.

Ben increases the enforcement costs of Dolly's diktats. It undermines the effectiveness of her attempts and increases the real costs of turning entrepreuneurs away from dangerous activities. This doesn't mean that regulations are never useful, but that regulatory policies imposed upon a decentralized system which is subject to perverse incentives is likely to be terribly complicated, expensive and inefficient. All sorts of tricks, from off-balance sheet accounting to off-shore corporate chartering.

The usefulness of exogenous regulations without policy moral hazard would be probably quite limited, and, besides, the entrepreneurial process is quite likely to be able to internalize many possible sources of "market failure". This implies that regulations may be mostly a solution (and quite often an additional problem) to a problem created by policymakers: if the political process were rational and efficient, it is likely that no such policies would exist.

However, it is so evident that policies often tend to be counterproductive, and that their inefficiency is quite never considered sufficient reason to avoid them, that it comes as no surprise that the best strategy is not often the chosen one. Even if it were intellectually evident that interventionist monetary policies are always counterproductive and that regulations wouldn't be necessary without them (I think these two statements are too strong, but not too much), it would be extremely unlikely that governments avoided the abuse of a policy instrument capable of reducing the cost of their debt, please voters by causing booms, postpone the redde rationem of economic restructuring when an unsustainable boom comes to its end, fund their activities, and give privileges to lobbies.

We have an aut aut. Either we have free markets that, because of policy moral hazard, will be rather dysfunctional and unstable, or we have a constrained market system, possibly with an evergrowing web of regulations imposed upon it, which will cause inefficiency in terms of foregone output and growth. We could solve the problem, or at least reduce it very much, by restablishing entrepreneurial responsibility by removing countercyclical policies, and this would likely reduce the usefulness of external regulations (endogeneous, i.e., entrepreneurial, regulations is led astray by policy moral hazard and thus cannot be trusted upon in these conditions); or we can keep on having the present system, but it is unlikely to be beneficial to society as a whole, other than politicians and lobbyists.

I think there must be some sort of a universal tendency in politics to hamper the working of decentralized institutions in order to substitute political subordination to market coordination. Cost socialization is the weapon of choice to convince everybody to evolve power toward the ruling elite in order to avoid the dysfunctionalities of a market process turned tragedy of the commons.

sabato 20 febbraio 2010

Financial regulation and monetary policy - (2)

Proven that decentralized decision making (e.g., the market process) can be lead astray by policies socializing costs and risks, because entrepreneurs will start exploiting the fake profit opportunities created by policy safety nets, and the entrepreneurial process of internalizing external costs is hampered by the possibility of exploiting the safety net, we are left with the option of regulating at least the most absurd aspects of market risk taking, for instance 30:1 financial leverage or 120 months/ 3 months maturity mismatch (mortgages vs abcp).

However, there are several problems with this approach.

First of all, regulations will both limit the entrepreneurial capacity of exploiting fake profit opportunities, thus increasing efficiency and lowering instability (moral hazard involves excessive risk taking, so limiting its effects involves increasing stability), and of exploiting real profit opportunities, thus hampering efficiency and growth.

Secondly, regulations will create an awful lot of public choice problems, which, considering that politics is rarely even close to be rational and efficient, will be usually beneficial to those in power and to organized pressure groups, but more rarely beneficial to society as a whole: I consider a matter of fact that few or no policies can be explained by assuming rational and benevolent politicians. In the field of monetary policies, given that the Fed in one century of existence has caused the Great Depression, the Great Inflation and now the Great Delevaring, I hold my tenet to be even more evident.

Thirdly, regulators can only use market-data conditional on other policies and regulations: risk prices distorted by moral hazard are all that a regulator can use to constrain risk-taking. As every market agent, policymakers included, cannot know more about the global state of the economy than prices tell them, distorted prices are a problem for the entrepreneur (which may be induced in error) and the regulator alike. It shall not be expected that the information destroyed by monetary policy interventions is available to regulators (and entrepreneurs): regulators are as blind as any other market participants when the price system is systematically distorted. This cost of hampering the market process will never be countervailed by regulations.

The first bottom line is that eliminating the source of moral hazard will enable the market process to work more stably and efficiently, and that no real alternative exists when the use of the market process is necessary to find and exploit information and to coordinate complex production plans. The intellectual division of labor enabled by the price system is effective as long as entrepreneurs foot their own bills. That's not what has been happening in financial markets, at least since Greenspan started his policy of systematic bailing out of those markets, if not in 1987, at least in 1990.

The second bottom line is that a systematic network of cost-socializing institutions and policies will hamper the market process and will result in squandering of resources and unproductive instability (this has nothing to do, contrary to Cowen's theory in "risk and business cycles", with choosing a point on the risk-return efficient frontier: moral hazard will enable additional risk taking even if it is not balanced by higher growth, so I'm talking of something completely different).

Bottom lines are, however, generally too clear cut. In Mises's terms, theory is accurate, often simple, and of universal validity, but history doesn't share these properties, and economic policies, like economic history and entrepreneurship, requires judgement and is never only a matter of pure theory. Thus, in the real world there are fuzzier cases than "the pure unhampered market process with no non-internalized relevant externalities" and "economic policies which systematically fail because of the knowledge problem", and, the causes of economic phenomena being never visible, the problem of causal imputation is never a theoretical problem, and often an untractable problem (in plain english: you can know that A causes C and that B causes C, but when you observe C, it's usually very difficult to say how much of the effect is due to A and how much to B).

What happens when markets have some element of inherent instability, other than that resulting from monetary policies and other sources of coordination failures among market participants? Can safety nets be justified? And can regulations be justified?

And, given that some safety net exists, is there a "second best" theory of optimal regulations that enable to smooth at least some of the problems created by moral hazard (or similar problems) which is not bound to failure because of some sort of a knowledge problem?

For what concerns the former objection, I would say that the existence of some tendency to instability inherent in the market process does not imply that market agents shall be shielded from the results of their own actions by safety nets. Unruly children are not made less unruly by overprotective parents: on the contrary, the more they smash their head against the floor, the less they will be likely to do stupid things. More seriously, to counteract the effects of some potential inherent instability by removing the private costs of bad decision making adds to the problem: it eliminates the only force that can internalize the potential sources of inefficiency, by removing the symptoms of the disease.

This is not, however, an argument against regulations limiting the possibilities of action of market participants, but only against safety nets: unruly children which are impeded to do stupid things may really cause less harm to themselves.


venerdì 19 febbraio 2010

Financial regulation and monetary policy - (1)

A week ago or so I wrote a comment on Thinkmarkets about regulations. The comment, as usual, was too long not to be boring and too short to be clear. As on my blog I don't mind being boring, I'll try at least to be clearer.

This was the original O'Driscoll's post, this my comment and this is an interesting paper by O'Driscoll regarding market-based regulations, which probably helps understanding the post and the following flame about bananas (banana is a technical term: check this comment). What I'm going to say can be partly found in this conference paper, although that paper focuses on a slightly different issue.

Financial regulation is a technical issue and as all technical issues is... ridden with technicalities. However, behind the intricacies of real markets and real attempts to steer and constrain them, there are a few theoretical principles of wide applicability which can be analyzed abstracting from the details. This is a result of the dicothomy between theory and history stressed by Mises.

This first post is largely introductory, and somewhat obvious, at least for those who have been exposed to some Austrian economics.

Decentralized decision making

The first point to notice is that on average a market works because profits and losses make bad decision making costly. This is not an obvious point: profits and losses depend on prices and prices shall depend on some underlying economic reality in order to make market coordination meaningful. For instance, in a lake subject to a tragedy of the commons, the price of using the lake is zero, the price of replenishing the lake with fishes is zero, and thus no one ever replenishes the lake and everybody fishes as long as the other costs of fishing (i.e., the fishing rod) are lower than the price of fishes.

In this case market coordination is not a free lunch and some redefinition of the structure of the game is required in order to make decentralized decisionmaking efficient: this process of restructuring the market is a form of entrepreneurship and constraints on entrepreneurship (i.e., transaction costs) stand in the way of this process of internalizing costs and benefits. The alternative to entrepreneurship is policy regulation, which however raises some important problems.

To summarize: the market process does not always work on autopilot (i.e., private optimality is not necessarily social optimality), but some assumptions are required for its smooth working, for instance the absence of relevant externalities not internalized by the entrepreneurial process. This process of "internalization" is part of the market process itself (at least as understood by Austrian economists), and it is one of the most important aspect of the market process.

Policy-induced "market failures"

The second point to notice, tightly related to the previous one, is that there are policies which create tragedies of the commons (or, more in general, "market failures": a term which is quite strange because it means "failure of the assumptions of standard neoclassical economics", which is tantamount to consider the existence of air friction as a failure of the universe instead of elementary mechanics).

Examples abound. For instance, if the cost of deposit insurance were not risk-weighted, or if risk weights do not adequately reflect true market risks (and how could they? Bureaucrats are not omniscient, and cannot recour to the market process without distorting it too much to learn from it), there will be profit opportunities which will not be exploited and false profit opportunities that will be created. Underpriced risks are a form of moral hazard, and will induce inefficient risk taking.

It is more seldom explicitly recognized that also monetary policy has a moral hazard effect, even without bailouts, recapitalizations, deposit insurance, monetization of private assets, credit guarantees by governments and government agencies (such as Fannie and Freddie)... the reason is that monetary injections have distributional effects: what I could have bought with my old money will be bought by someone else with his new money. The result is that when something is bought with new money, the costs are shared by everyone has to face (counterfactually) higher prices. For instance, if the entrepreneur who gets a loan from the bank makes the price of machinery rise, it will harm all other entrepreneurs.

Under these conditions, I'm rather surprised things work

At this point, a question must be asked: how can anyone think that the market process can ever work under widespread, systematic and long-term monetary policies which create this cost socialization problem? If entrepreneurs can buy and sell goods by shifting part of the costs on other entrepreneurs, on final money-holders, or on other market agents (for instance, bond-holders receiving lower returns, tax-payers which have to refinance banks, etcetera), how can this decentralized market process lead to socially beneficial results?

The standard answer, of course, is to regulate. However, there are several problems with this approach.


lunedì 15 febbraio 2010

Hello world

This blog is supposed to be my notepad about Economics. I'm not an economist and so you shouldn't expect this to be a professional blog. Besides, even if I were an economist, I couldn't write anything that could be used to pass an exam, because I'm mostly interested in Austrian Economics.

I'm going to link blog posts, newspaper articles and academic papers I like, and occasionally write a comment. I'm mostly interested in macroeconomic issues and so most of the posts will be related to the financial crisis and other not-so-funny topics.

Comments are free (although this blog will be so boring that I don't expect any), but I'm not interested in flames, and easily bored by prolixity: I prefer suggestions and analyses on the economics literature. That's also the reason I'm not going to post what I don't like: I prefer to be constructive.

Probably the greatest problem with Austrian Economics today is that it is the most famous economic theory on the blogosphere, which is not something to be proud of, considering the low average quality of web discussions. I'm interested in the wheat and not in the chaff, and I hope to be able to distinguish them.

I created this blog because I often comment on other blogs, but as I most often write when I have a doubt or when I disagree, I fear to appear overly critical. Besides, not to write too much I often write too manicheist comments. I hope blog posts will improve upon my comments.

I hope this blog will be more original than the title of my first post.