Brad DeLong's Semi-Daily Journal (2004)

Please explain something I don't understand. I thought that efficient markets had to be large and have many actors, and that small countries were too small to have a large market in anything.
The idea that the US model with the huge currency, bond, equity, labor, land, and commodity markets we have could be applied to some dinky country appears to be absurd.
When you consider the way the US stock market yo-yos around, think what it must be like for even a 'middle' sized country like Argentina.
So why is this arguement even taking place?

Read the comments from those posts. A bunch of intelligent people got into those arguments, not the least of whom being James Surowiecki, who just wrote a book on this whole topic.

I really do agree with D-squared's critique against efficient markets. But in developing an emerging market, or industrializing a country, it has to be considered that markets only allocate small chunks of money at a time. In an underdeveloped country without industrial infrastructure, small-scale (firmsize) investment can probably not pay off enough to feed its workers. So the market does not allocate any capital to it.

However, an investment large enough to build a meaningful infrastructure could pay off, given low but still liveable wages. It might even be optimal. But market participants would only do so if they believe the rest of the market to allocate enough capital to create enough benefits of scale (i.e. infrastruct.). Hence the kind of extreme emerging markets volatility that D^2 is referring to.

And a global planner that could allocate nation building amount of capital would have a chance to outperform the planner.

Still - it has to be nothing wrong with the market's allocation of the next few dollars. And to me it seems that we recently have had an inflation in financial market disbelief. "Look, the dollar is soooo overvalued!" "Bubble!" "Look, gov't bonds!" etc... Haven't we waited long enough on these "bubbles" to pop to claim an explanation from the experts whome have been crying "wolf" for so long...

Prof Brad Delong and the Academic Left still don't understand economics...
And here's why:

All their arguments against the Efficient Markets Hypothesis in favor of government planning are not only wrong but miss the point entirely.

The genius of free markets is that they allow evolution by natural selection to occur as a matter of course in the economy. Nations that replace free markets by central planning can never really evolve - get better - because the planners won't allow it. This is why the USA is so much richer and more successful than the Eurosocialist democracies. As evidence I assert that the average American earns $9,000 more per year and lives in twice the space of the average Social Democrat citizen.

The main mechanism that allows this is called DIVERSITY in politics and biology and DIVERSIFICATION in finance. They mean the same thing and work the same way in their diferent domains. It has to do with putting all yopur eggs in one basket.

Democrats get DIVERSITY all wrong because...

http://pep.typepad.com/

Adrian --

I don't think that anybody, from Brad to D**squared is in favor of a completely planned economy. They mostly seem to be in favor of Krugman-in-east-asia-style interventions and controls. Whether or not that can be turned into a coherent system is a different question, but your argument that since a completely planned economy fails, a completely free market must be the best solution, *really* misses the point.

Just to show I am not a sinlge note, there has been some recent research on converting the value Europeans get from their expanded social services, longer vacations, better medical care (perscribed spa treatments perhaps?) that narrow the earnings gap considerably. Personally I would take the $9 grand but hey it seems to work for them. If there are some that don't like it they can take the options our ancestors (or atleast alot of them did) and leave.

I think the discussion was more about developing countries that are or by choice have been insulated from larger and somewhat more efficient world markets and are to small to support any reasonable internal markets. I am sure it was an interesting discussion, where there any general conclusions reached other than there are no such things as free lunches?

Adrian Spidle wrote: Prof Brad Delong and the Academic Left still don't understand economics...

Man, you are so funny. Can you wiggle your ears as well?

Adrian, you're confusing informational efficiency with allocational efficiency. They aren't the same thing.

The key to understanding distributed processing events is to understand that they are prone to being biased by systematic biases within their network participants. For instance, racism can occur as a systemic bias even though individual participants don't overtly practice racism, simply because of an implicit coordination of standards and perceptual assumptions.

In the same way, markets are no more efficient than the regulatory structures of a market and the dynamic of market participants allows it to be. All markets have an ideal market efficiency based upon the regulatory structures and how the dynamic of the participants interacts with those. the regulatory structures may be formal (legislative) or informal (traditional or negotiated) but nonetheless they exist.

I'm working on a game theory algorithm for monte carlo and game theory decision making that should hopefully be able to simulate the efficiency of market participants exchanging information and transactions for a given regulatory structure. Of course I'm also looking for a new career while I do it, so it's getting done in my less than copious free time.

Sry, I got off the informational track as well. Still fascinated by Pentagon trial balloon about Terrorism market, hehe. Maybe they should just call the psychic friends network.

Anyone even considering taking Adrian Spidle even partially seriously is advised to check his blogsite that he is so proud of trumpeting. Really scary stuff, sort of Ayn Rand meets L. Ron Hubbard. I'm actually glad I don't understand what he is talking about.

Umm. Can we please be a bit more clear about a few terms. The debate about the 'efficient markets' hypothesis is NOT a debate about whether decentralized markets are better or worse than centralized planning.

The EMH debate is about whether there are any arbitrage opportunities to be found and exploited in financial market for secondary assets such as stocks and bonds.

Many economists are perfectly comfortable saying that they do NOT believe there are any exploitable arbitrage opportunities on such markets and hence that they believe that the EHM applies, YET they may feel perfectly comfortable arguing that market failures are rampant in the world of finance (as someone above argued, how else can you explain firms or financial intermediaries such as banks).

Think of it this way. Do you think there are many arbitrage opportunities in the world of mortgaged back securities? Highly unlikely given how much competition there is to add such securities to portfolios in this multi-trillion dollar industry. Now ask yourself whether this market would be even one fourth as large as it is today had it not been for government interventions in the form of Fannie Mae and Freddie Mac to help develop those markets (whether these institutions are now too big or have outlasted their purpose is a separate question).

Morgaged back securities may trade efficiently, but banks will always ask for collateral. The last observation is irrefutable proof that markets are not efficient.

Oldman, that sounds like a heck of a problem. The interesting ones always are. I think if you can build an effective algorithm, the world will beat a path to your door (atleaset academic world). Truly best of luck both in that and in new career search.

Non-economist who would you nominate as the secrect love child of Ayn Rand and L Ron? A scary thought indeed. To match the parents the child would have to create a non-totalitarian self help cult of personality.

Umm. Can we please be a bit more clear about a few terms. The debate about the 'efficient markets' hypothesis is NOT a debate about whether decentralized markets are better or worse than centralized planning.

The EMH debate is about whether there are any arbitrage opportunities to be found and exploited in financial market for secondary assets such as stocks and bonds.

Many economists are perfectly comfortable saying that they do NOT believe there are any exploitable arbitrage opportunities on such markets and hence that they believe that the EHM applies, YET they may feel perfectly comfortable arguing that market failures are rampant in the world of finance (as someone above argued, how else can you explain firms or financial intermediaries such as banks).

Think of it this way. Do you think there are many arbitrage opportunities in the world of mortgaged back securities? Highly unlikely given how much competition there is to add such securities to portfolios in this multi-trillion dollar industry. Now ask yourself whether this market would be even one fourth as large as it is today had it not been for government interventions in the form of Fannie Mae and Freddie Mac to help develop those markets (whether these institutions are now too big or have outlasted their purpose is a separate question).

Morgaged back securities may trade efficiently, but banks will always ask for collateral. The last observation is pretty powerful proof that markets fail.

You can't have markets without enforcement, yet enforcement is pretty difficult without states (go ahead and argue that enforcement is possible without states, and I will agree, but I'll also point to how such contexts soon lead to the development of institutions and states).

There are a lot of markets for instance Jonathan such as commodity markets, options markets, IPO issuances, Ebay, real estate, etc. which have profound economic significance but are as important than the secondary market for equities you cite but efficiency is a concern. For instance take the market for sovereign debt, if there was a market whose efficiency depended upon regulatory quirks and governing policy there was one right there. Market efficiency hypothesis needs to be more sophisticated to deal with these situations.

Adrian Spidle said:

"Nations that replace free markets by central planning can never really evolve - get better - because the planners won't allow it. This is why the USA is so much richer and more successful than the Eurosocialist democracies."

How much richer is the US and, more importantly, by which measure? Popular economic indicators are (1) calculated in a different way, and (2) compared out of context.

"As evidence I assert that the average American earns $9,000 more per year and lives in twice the space of the average Social Democrat citizen."

I would submit, not assert, evidence, but anyway.

This is presumably meant to mean "the average US wage is $9000 higher than the average European wage, adjusted by the respective exchange rates". Aside from whether that's accurate, it misses other important determinants of living standards, like unemployment, social security, welfare, disability benefits, effective tax rates, price levels, and a host of other social policies. On those I think it is fair to say that Western/Central Europe looks far more favorable than the US.

Taking the interpretational liberty to replace "Social Democrat" by "Old Europe", yes, after having moved to Califonia I have been making not $9000 more, but _2x_ as much as in Europe (lets keep this in relative terms). I pay slightly lower tax (income/payroll) + healthcare insurance rates, but I pay more than 3x for accomodation, 2x for food, 0.5x for gasoline, about the same for non-healthcare services (hard to compare), and I leave a 4-digit amount at dentists alone almost every year (thank the Lord I don't need to go to doctors yet)!

And looking around, in Germany we (used to) have relatively good welfare/pension/unemployment benefits and quite close to universal healthcare, which accounted for a good part of my taxes, perhaps 20+ percentage points. When adjusting for that, my effective tax rate in Germany looks much more favorable than my US one.

As for living in twice as much space, considering the far greater landmass of the US it does not look like that much of an achievement. As roads have generally more lanes in the US due to more space and more commuting (in twice as large cars?), I figure even the homeless can benefit from broader bridges and end up living in twice as much space.

DIVERSIFY that!

Oldman: Just out of curiosity, any idea yet what software/language you would use for your research?

Umm. Can we please be a bit more clear about a few terms. The debate about the 'efficient markets' hypothesis is NOT a debate about whether decentralized markets are better or worse than centralized planning.

The EMH debate is about whether there are any arbitrage opportunities to be found and exploited in financial market for secondary assets such as stocks and bonds.

Many economists are perfectly comfortable saying that they do NOT believe there are any exploitable arbitrage opportunities on such markets and hence that they believe that the EHM applies, YET they may feel perfectly comfortable arguing that market failures are rampant in the world of finance (as someone above argued, how else can you explain firms or financial intermediaries such as banks).

Jonathan, this is another version of what it pleases me to call the "Surowiecki gambit" which we discuss a bit in JQ's post. The unpalatable truth is that these two debates are one and the same. If there are no arbitrage opportunities, then prices fluctuate randomly, and (per Samuelson) we know that prices incorporate all available information about the future. If prices do not fluctuate randomly, then there is additional information about the future which is not incorporated in market prices, and there is therefore an opportunity (perhaps only a small opportunity) for a planner to improve on the market process. An arbitrage opportunity (or even a deviation from fair value which is not an arbitrage opportunity) is by the same token, a case in which someone using full information could outperform a market which is not using full information.

Pleased as I am to have a gambit named after me, here's the way I'd put it: even in the best -- most liquid, most information-rich -- asset markets, the price of any individual asset rarely, if ever, incorporates all of the available information about the future performance of that asset. It's therefore possible, in principle, for a planner who had access to all the information that is incorporated in the asset price and all the information that isn't to do a better job of setting an accurate price for that asset. (An accurate price here is assumed to be a price that reflects the NPV of all its future free cash flows.)

There's also no doubt that when it comes to the price of any single asset, there are almost always investors -- and therefore there could almost always be planners -- who have a better picture of the future than the market does. The first question is whether, over the full array of assets that the market has to make decisions about, any investor or planner can consistently offer better forecasts than the market. The second question is whether we can identify that planner in advance. I think there are particular historical moments -- call them bubbles, or panics -- when this may be possible, and there are also particular kinds of markets where it's almost certainly possible. But most of the time, I doubt it.

The fact that there is available information about the future that is not fully incorporated into market prices does not imply mean that an individual planner would be able to consistently find and recognize the value of that information and incorporate it into his prices. To know if it would, we need a rigorous analysis of planners' performance -- not simply a recounting of those moments where the market was obviously wrong and an individual planner happened to get it right. Give us a real Efficient Planners Hypothesis, and we'll be able to find out how well it works.

James Surowiecki points to the question of whether the market will choose the best equilibria when there might be multiple equilibria, and whether the state can help markets coordinate on the 'right' one (e.g. help the market arrive at a 'focal point' or help deflate a bubble). Here again is an issue where the believers in EMH can point to a bunch of technical arguments (for instance as to why there really should not really be multiple equilibria or bubbles) whereas those on the other side can point to a real world institution that seems to exist to prove their point: Alan Greenspan.

oldman writes:

"The key to understanding distributed processing events is to understand that they are prone to being biased by systematic biases within their network participants. For instance, racism can occur as a systemic bias even though individual participants don't overtly practice racism, simply because of an implicit coordination of standards and perceptual assumptions."

Well, yes, but systematic biases are A) not held by every market participant and B) are up for review at every transaction. The regulatory side has the same proclivity for bias, but throws you in jail for disagreement and does not self review. Public Choice matters, too. The interface of politics and economy is frikin' ugly. A market connects them, but it is one that rewards political clout more than willingness to pay.

That aside, I'm glad to see James S. note above that you have to count all the times a given planner is wrong. We are largely victims of the principle of Odd Matches, where we remember only the times the planner is right and the market is wrong, forgetting in the process that we note this event as memorable precisely because it is an exception.

Brad,

This people are all well intended, but miss the mark, badly. The important information is not about the market, it is about externalities to the market.

The reason there are firms is that there are people with better skills, insights and ability than others. Second, not all the important information is transmitted by markets. Three examples.

EX 1. Everyone has heard the story about the investor who bought Wrigley's when he noticed the gum and gone up in price from 5 cents to 10 cents. This was information about one market, but it took individual skill and insight to use such in a second market.

Ex 2. In Noble House, the wealthy lived at the top of the mountains over Hong Kong harbor. This permitted them to know sooner what ships were approaching and, thus, when to sell before new supplies drove down prices. This was information about a market, now transmitted by a market.

EX 3. Lance Armstrong won this 6th Tour for several reasons, but principally because he had better information about: (1) his own health and fitness; and (2) the roads and how to ride them. This information could never come from a market.

The living, breathing proof of all this is Warren Buffett. Please read what he says about "efficient markets" and such other silly stuff.

Best regards,

Adam Smith

I had a class in b-school with Richard Thaler, probably the most prominent figure in "behavioral finance," the movement most often identified as a serious challenger to the EMH orthodoxy that has reigned in academic finance for a quarter century.

The behavioral finance people will point out all sorts of ways that markets are sub-optimal because of systematic biases in the way people make deicisions - people are myopic, they are over-confident, they don't equally weight value-equivalent gains and losses, etc. But on the last day of class someone asked Thaler a question, the tone and substance of which implied that since he had spent so much time knocking down the intellectual underpinings of EMH, that he thereore must believe in an expansive role for governments in regulating the economy.

Not so, he declared emphatically! Rather, he pointed out, the same quirks of judgement and decision making that impair actors in markets also impair actors in governments.

Just because you identify an instance in which the market acts stupidly doesn't mean that you'd get a better outcome if you had stripped the market of power over that instance and replaced market forces with the judgement of a government planner. The planner, after all, is just as human as the people who make the market.

Now, anyone with a grounding in reality (by which I mean most people but not, say, academic Marxists and Atlas Shrugged cultists) will agree that sometimes government intervention in the economy is good, and sometimes its bad. Reasonable people differ on the proportion of the two cases. I myself tend to be a market booster - I'm suspicious, very suspicious, of the ability of centralized authorities to add value in most circumstances. But I'm happy to argue the merits on a case by case basis.

One other point, in collegial opposition to Mr. Davies. The point has been made on several of the Crooked Timber threads that most capital is not, in fact, allocated by markets, but is allocated within firms by small bodies of experts with centralized authority. True indeed.

But Mr. Davies seems to imply that this is a good thing - that such capital allocation decisions are better than the decisions made by markets. To which I must collegially reply: Are you fuking smoking crack? I work with large corporations every day. And I can assure you that you'll win no arguments by pointing out the virtues of how companies allocate their capital. Internal capital allocations schemes are stunningly stupid in a majority of cases. I've seen, time and time again, companies in dire financial straights because they poured money into existing, often outmoded, lines of business like it was water and refused to commit resources to new opportunities if those opportunities couldn't meet a 20% hurdle rate (applied, illogically, to cash flow projections that were already probability-weighted).

Markets are often criticized for bubble behavior - which is typically just an irrationally large appetite for risk. But non-market capital allocation systems have an often irrationally large aversion to risk. The stock market will give $20B to Webvan.com. Stupid indeed. But the "experts" running corporate finance at Old Economy Grocers, Inc. will put a roadbloack on investing in IT infrastructure projects that lower operating costs enough to pay for themselves 5 times over in the first three years because self-important sense of fiduciary duty drives them to demand of anything new under the sun far, far more than they demand from what is old and tired. Also, IMHO, stupid indeed.

In discussing EMH, "consistency" of beating the markets is an issue (e.g. in James S. writes about a planner who can consistently offer better forecasts). Does anyone know if there is an agreed on understanding of that term? The term strikes me as somewhat vague. Can't expect someone to be right every time on every decision. But if not, then how often, and over what time period, do you have to be right to be "consistently" better?

Actually, Walons, for someone who believes in EMH, consistency of beating the market isn't an issue, because there is no way to do it -- even on a single asset -- unless you're an insider. Now, I don't think anyone really believes that anymore, but the idea of "consistency" in the way I use it hasn't really been that well developed in the academic literature. I haven't formulated a rigorous definition, but I would say a commonsensical definition would be something like this: over a period of time and a number of decisions that seem large enough to be statistically meaningful, if a planner is more accurate more often than the market, then I'd say he or she is better.

The metaphor I use in my head is the racetrack. If over, say, 100 races, a bettor can do a better job of forecasting the odds on horses winning, then I'd agree that he's smarter than the collective judgment of the crowd. But one thing to keep in mind: to be a fair test, the bettor has to not just pick winners. He has to offer odds on every single horse in every single race, since that's what the crowd at the track does via the parimutuel pool.

I'm also very much with SD in his critique of the way firms allocate capital.

Btw, I'm using the term "planner" in order to maintain full generality, but that brings in the unfortunate connotation of a government planner. Note that for most cases (I think everyone on this thread is clear on this but just to make it explicit) the planners for a capital-allocating unit are the managers of a business.

Which in turn sheds a bit of light on James' claim. Another prediction which was derived from the EMH (this one is mainly due to Prof. Michael Jensen) was that companies should pay out most of their free cash flow as dividends (or share buybacks). The idea here was specifically the one at the heart of this debate; the companies had a choice between two things to do with their cashflow:

1) reinvest it themselves, leaving the decision about capital allocation to insiders
2) pay it back to shareholders who would reinvest it on the market, effectively putting the capital allocation decision out to the market.

Jensen's idea was that 2) was clearly a better way for capital allocations to be made, because markets were efficient, and that therefore the observable fact that companies did not pay out much of their free cash flow had to be explained by some pathological outcome of companies as organisations. This was the intellectual underpinning for what came to be known as the "leveraged buyout boom of the 1980s".

It was, strange to say, a disaster. The LBO boom didn't bring any great improvements in capital allocation, despite benefiting from a massive public subsidy in the form of deductibility of debt interest. I maintain that the empirical evidence strongly supports the view that the problem of planning capital investments, or the second-order problem of finding people who can plan capital investments, is nowhere near as herculean a task as people suggest, and that markets don't actually reproduce an expert's judegment on the question particularly well.

DD

Please post all your notes on EMH in sequence. They are most interesting and useful, and I wish to make sure I have them all.

Now, I don't think anyone really believes that anymore, but the idea of "consistency" in the way I use it hasn't really been that well developed in the academic literature

Not wanting to cast aspersions or anything, but it has, man, it absolutely has, in the literature starting with Wald on optimal experimentation, going through a whole load of Bayesian statistics and ending up probably with Spyros Skouras and "decisionmetrics".

Btw, I'd also point out that racecourse bookies do exactly what James is talking about above, and appear to do so with a decent degree of success (though it is not always clear to me the extent to which they are acting as market-maker)

James/DSquared,

I guess my problem is not that there's been no attempts to define consistency, but rather that no one definition seems to have gained universal acceptance, in the sense that it's both statistically rigorous and agrees with our intuitive preconceptions. For example, presence of long failure runs don't seem to agree with consistency, even if the overall expected (excess) payout is positive (Buffet).

Interestingly, I'm actually writing a piece on this right now. What's the "empirical evidence" that companies do a great job of allocating capital? I would say that the simple fact that 2/3 of all mergers -- the single most important capital-allocation decision in terms of actual dollars, and one typically made by the CEO -- destroy shareholder value should make one seriously skeptical about trusting a single individual to make planning decisions. I would also say that the disastrous outcome of the conglomerate boom of the 1960s -- entirely the product of individual planners believing they could do better than the market in diversifying -- casts doubt on your thesis. But regardless, what's the evidence on the other side?

In any case, I think firms could do a far better job of making capital-allocation decisions by relying on a wider pool of decisionmakers rather than simply on the "experts" at the top. That's not the same as relying on the market, but then no one I know has advocated letting the market decide whether, say, Microsoft should invest in one product rather than another. No one -- least of all Michael Jensen -- would deny that the people at Microsoft are in a privileged position to evaluate their portfolio of possible investments, and that given the right circumstances they're more likely to choose the best ones. The question is how often do those right circumstances prevail.

Jensen's argument was not as crude as saying that markets are efficient, firms are not, so give the money back to the market. He was saying that in firms that were flush with cash, where the incentives to make value-maximizing investments were small, managers were likely to use the money irresponsibly -- most notably, by spending it on wasteful takeovers. There are a pile of studies showing that this is in fact more likely to happen in cash-rich firms, especially those that have just become unusually flush.

As for the LBO boom, I'm not sure what you mean when you say it was a disaster. But in any case, if it was, it's an argument against your position, since the people who ran LBO firms and did the acquisitions were almost prototypical planners -- they were a small number of experts who believed they could value companies better than the market (otherwise they wouldn't have been able to make huge profits taking companies private). It seems peculiar to say that taking a company private -- which is what the LBOs did -- means giving the market a larger role in making internal capital-allocation decisions. LBOs were intended to provide more of a check on companies by ladening them with debt, but it seems hard to say that that's any more of a market-driven process than, say, companies relying on bank loans to fund their operations.

Now, anyone with a grounding in reality (by which I mean most people but not, say, academic Marxists and Atlas Shrugged cultists)

Now, I'm not an academic Marxist, but I can tell you that virtually every neo-Marxist academic in the field of political economy I know of DOES NOT believe that government planning is a great thing all the time. The current bend of most theorists in this area is that the means of production should be under the direct democratic control of the public. Right or wrong, at least bother to accurately characterize the current view of academic Marxists, would you.

(In case you care, I'm mainly a neo-Gramscian when it comes to IPE, and no, I'm not a socialist of the direct democratic or state capitalist stripe, either).

At least in the US, bookies don't set odds on horses at the track. And in other sports, bookies try their best not to take a position on either side of the bet, so that they're simply making a market and clearing a steady profit that way.

Takeovers and mergers are completely atypical capital allocation decisions.

I've not spelt out my own view at length yet, but the key to it is my original "Hayekian markets reconsidered". People make good decisions when they are dealing with matters where they have tacit knowledge. People who only have propositional knowledge don't make good decisions, and getting a gang of them together and calling it a market helps less than you'd think.

I disagree that Jensen's argument was "not as crude as that"; in particular, have another look at some of his articles in the "Journal of Applied Corporate Finance", collected in Don Chew's book "The New Corporate Finance". Some of them are extraordinarily crude.

An LBO firm is one in which the capital allocation process is pre-empted by the contractual payments on debt. It is this pre-emption of the judgement of the parties wih tacit knowledge that I regard as dangerous, whether it comes about as a result of markets or otherwise. It's part of my general campaign to rehabilitate subjective judgement and my war on black boxes.

sd states, and I quote: "..Internal capital allocations schemes are stunningly stupid in a majority of cases."

Then markets must REALLY fail badly because last I checked internal capital allocation schemes and retained earnings accounted for the lion's share (I'm talking over 70%) of new capital formation in the industrialized world.

There's a variant of 'if you're so smart why aren't you rich' to be applied here. If internal capital markets and relationship banking supposedly fail so miserably compared to arms-length finance, then why the hell hasn't the latter taken over?

The point, again, is that EMH may very well apply to the trading of assets on _secondary_ markets, since this simply entails reshuffling very easily traded assets between investor porfolios on very thick markets. It's a complete non-sequitur to say that that proves that capital markets are efficient.

Most firms, except for the very largest ones do not use bond or stock markets to finance their ongoing operations or NEW capital expenditures. Why? Because they simply wouldn't be able to get the markets to purchase their IOUs, or not at rates that compete with internally generated funds or loans from banks. That indicates a number of market failures associated with information asymmetries and costly enforcement. Internal capital markets (i.e. firms) and banks exist because they solve these financing problems more efficiently than the sort of arms-length contracts that could be traded on secondary markets.

The EMH is therefore not relevant to the vast majority of financing situations in the world.

Government is highly unlikely to improve on private fiancial intermediaries in the provision of this kind of finance, but there is a load of room for policy, because policy profoundly affects the operation of banks and financial intermediaries.

Hence I return to my original claim that finding evidence of the operation of EMH on certain specific markets has very little to do with the debate over states versus markets.

DD

Please let me know how I can collect all your comments on this topic. Thank you so very much.

Daniel, in lots of firms takeovers and mergers are not completely atypical -- they're central to the way the firm grows. Even in firms where this isn't the case, which is certainly most companies, takeovers and mergers are usually far more consequential in terms of the amount of capital allocated and the impact on the firm than any other decision an executive makes. If individual planners do a consistently bad job of making those decisions -- as we know they do -- that has to be included in any evaluation of the performance of planners versus markets.

As far as the LBO question goes, I don't see how contractual payments of junk debt "pre-empt" the capital allocation process any more than making installment payments on bank loans do. They certainly constrain what the managers of the company can do, but by that standard any limits on the managers' potential actions (including their inability to raise more capital to fund projects) would count as market pre-emption.

In any case, I think you need to put the arguments for LBOs, and Jensen's critique of managerial authority, in historical context. After fifteen years of unconstrained American managers running American industry into the ground, building paper empires, and making value-destroying investments, there was reason to believe that imposing some external discipline on them would be beneficial.

More broadly, I think the line between tacit and propositional knowledge is not as well-defined as you suggest. And in any case, unless you think the people with propositional knowledge are all going to be biased in the same direction, getting a gang of people together -- some who will have tacit knowledge, some with propositional knowledge, and some who know nothing -- helps your decisionmaking enormously. I'd agree with you about subjective judgment -- I think collective decisionmaking processes work, when they do, because they aggregate subjective judgments. But that may be black-box thinking.

Back to an earlier question, though: what is the empirical evidence that firms do a good job of allocating capital internally?

Jonathan wrote:

"sd states, and I quote: "..Internal capital allocations schemes are stunningly stupid in a majority of cases."

Then markets must REALLY fail badly because last I checked internal capital allocation schemes and retained earnings accounted for the lion's share (I'm talking over 70%) of new capital formation in the industrialized world.

There's a variant of 'if you're so smart why aren't you rich' to be applied here. If internal capital markets and relationship banking supposedly fail so miserably compared to arms-length finance, then why the hell hasn't the latter taken over?"

The latter hasn't taken over because it is extremely costly to set up an arm's length market. It generally makes sense to let markets allocate capital only when the specific decisions are either very, very big or when the capital allocation can be efficiently liquidized. When a company invests in new machine tools for an assembly line, the decision is not big enough for a market assesment to be cost effective, and its an all-or-nothing decision (having 3/4 of a mchine tool is of no value), so it makes no sense to securitize the decision.

But that doesn't mean, as I took D^2 to be saying, that internal capital allocation schemes are better than markets at making good decisions. It simply means that markets are so difficult to establish that it makes sense in many cases to stick with a substantially worse method of allocating capital.

We wouldn't say, for example, that a Ford Taurus is a better car than a Volvo, though the former accounts for a much higher percentage of vehicles sold. Rather, we would say that you'd better off getting a Volvo (possibly even on an NPV basis) if you could afford it, but for most people the Ford is the only option.

As far as the specifics of the "excess cash" argument go, the Arnott/Asness paper showing that future corporate earnings growth is fastest when dividend payout ratios are high and slowest when the ratio is low at least suggests that getting excess cash out of the hands of managers and into the hands of shareholders is, in fact, a good thing for corporate performance.

DD: "these two debates are one and the same. If there are no arbitrage opportunities, then prices fluctuate randomly, and (per Samuelson) we know that prices incorporate all available information"

But again: there are different infornation sets - the market has to guess what investment others will make, a large enough planner makes has not, it makes most of the investments itself.

D2, as far as I can remember, Jensen's arguement didn't need EMH. Anytime a company becomes more highly levered it transfers money from bondholders to shareholders (since shareholders still have the upside risk, while bondholders now shoulder more of the downside risk).

I thought the whole topic was more about illustrating the various conflicts of interest presents by the various stakeholders (typically management, share holders, and bond holders). From the perspective of shareholders companies are not levered enough, merge too often, CEO's pay is too high, have free case etc. From the point of view of management however, ...

As James points out, these things are really not market failures so much as planner (management) failures. I'm sure there are a number of people out there who don't feel they are failures at all :).

Lorenzo-Since I don't often come into contact with academia on these subjects could you clarify what the practical implication of "means of production should be under the direct democratic control of the public"? I assume that would mean the democratically elected state would appoint a manager to run the means of production, which does not seem a lot different than centralized government planning. Perhaps I am being obtuse.

sd,
I had to laugh at your comments about the stupidity of internal corporate allocation methods. I think that outsiders would be surprised to see how things actually work. Several years ago I was being taught how to do NPV analysis by an old corporate guy. He said, "Well, the first thing I do is divide whatever Marketing forecasts in half." This obviously wasn't policy, but at least he kept some silliness at bay.

>>As far as the LBO question goes, I don't see how contractual payments of junk debt "pre-empt" the capital allocation process any more than making installment payments on bank loans do.

You might want to ponder on what it is that the "L" of "LBO" stands for ...

Note that almost all the empirical work on takeovers focuses on takeovers of one publicly traded company by another. This is quite unusual as a capital allocation decision because it actually requires consultation of the capital markets (both to win the proxy battle and to finance the issue of debt and equity which is usually part of the financing).

I'm afraid that I don't regard the Arnott/Asness paper as showing anything because, IMO, it's not a very good piece of work. In particular, the dataset they used (in the Financial Analyst's Journal piece anyway; I think there is another one) is a real Heath Robinson affair cobbled together from stock returns and earnings yields and there was no real discussion of how they might have dealt with a number of serious problems you could introduce in this way. I also thought that their attempts to check whether they were picking up simple mean reversion in earnings were pretty statistically naive.

In answer to your question, I'll simply rely on the wisdom of crowds. I don't believe that markets are good at making specific decisions on specific investment projects, but I do think that they are good at doing what they're there for; valuing traded securities. The average book-to-market ratio of US traded equities is less than one and has been for a long time. This only makes sense if the firms are, on average, investing their capital at a rate of return higher than its opportunity cost. At present, Tobin's q-ratio is higher than 1, which suggests that firms are investing at a higher rate of return than the *marginal* opportunity cost.

Dex,

Lorenzo-Since I don't often come into contact with academia on these subjects could you clarify what the practical implication of "means of production should be under the direct democratic control of the public"? I assume that would mean the democratically elected state would appoint a manager to run the means of production, which does not seem a lot different than centralized government planning. Perhaps I am being obtuse.

The key part of the phrase is "direct," meaning direct democracy rather than representative, usually at the local level. Thus it would be those who produced any given product who made all the decisions under a rubric of decentralization and localized direct participatory decision making. Perhaps the closest existing example would be non-profits and co-ops where those running the firm decide what to produce, and how.

Among advocates of socialism these days, it is generally thought that doing everything at the state level is impractical and too open to creating a cadre class of beuraucratic elites and the the further decision making can be decentralized to the local level and done in a direct participatory fashion, rather than through a representative system, the better.

That is as best as I can make out their arguments, anyway.

Daniel, I'll agree that acquisitions are unusual. But to paraphrase what you said in a CT thread to the person who was talking about Argentina's currency policy not being market-driven, no one is forcing CEOs to make these acquisitions, unusual or not. If the stock market's valuation of Pets.com is an example of the idiocy of markets, then AT&T's acquisition of NCR for $110 a share, or Bernie Ebbers' expenditure of $50 billion on a plethora of overpriced telecom companies are examples of the idiocy of planners. If we're trying to evaluate how likely it is that an individual planner, so to speak, will make good capital-allocation decisions (or more broadly, good forecasts) then the miserable M&A track record of CEOs is absolutely relevant.

In any case, I sometimes have the feeling -- which I had when we were discussing this in the Jane Galt thread -- that you think I'm making an argument that I'm not making. In the context of firms, my critique of the CEO isn't intended to be an argument for letting the stock market make specific decisions on specific investment projects. As you know from the book, it's an argument for changing the decision-making process within firms, so that instead of the CEO having the final say, the decision reflects the collective judgment of a much larger pool of decisionmakers. The exact process by which the aggregation of those decisionmakers' individual judgments takes place isn't, I think, especially important, as long as the judgment is really a collective one. Now, you can critique this by saying that an individual CEO will still outperform a crowd of executives, but I think the evidence really demonstrates that he won't.

Finally, with regard to the internal allocation of capital question (which is really what the CEO discussion is about), I don't see how the fact that companies are, on average, doing a good job of investing their capital means that they couldn't be doing a significantly better job. That "on average" could be -- I believe is -- hiding a lot.

Brad, care to give us a hint as to what you have in mind - regarding "an integrated theory of market *and* government *and* regulatory failure". What will it look like? Whose work will it build on?

This is a great thead-- I'm learning a lot.

For my 2 cents. I agree with the argument that corporations do a bad job of allocating capital.
No corporate vice president worth his salt ever proposed an investment that did not guarantee at least a 33% rate of return. But they get to that point backwards. If they want make an investment that make whatever assumptions about demand and prices will be when the investment comes on stream to generate the rate of return needed to justify the investment. I will never forget that as an analysts going to company presentation after company presentation in the early 1980s and hearing company after company say the high rates were not imacting their investment programs. But on the other hand the way the market allocates capital to the economic sectors that need it is to allow them to make superior profits -- most likely because there is inadequate capital in that area . So the firms in the sectors needing new investment get the cash flow to spend on adding capacity in that industry.

We are seeing that now. While we overinvested in tech in the 1990s we underinvested in basic industries. Now we have surplus capital in tech and shortages of basics. So commody prices rise and basics make more money and spend the money to expand their capacity while tech stagnates. This highly inefficient way of overshooting and undershooting is the way markets really work--
it is highly inefficient and the theory that it is efficient is a bunch of bull.

Meanwhile those arguing that govt or planners are bad use the example of the communists and ignore Japan and the Asian tigers. Govt planners played a much larger role in allocating capital than free markets over the last 50 years and they have had some of the greatest economic success stories in history.

The argument is not between the 2 extremes, reality is along a spectrum of both public and private and sometimes one is better and sometimes the other is. But no one can predict ahead of time which one.

anyone who can look at what happened in the bubble and still argue that capital markets are "efficient" is not being honest.

Just what are they supposed to be efficient at? Stripping wealth from the majority, concentrating power, ending competition thru bigger and bigger corporate structures?

Just to add belatedly my $.005 worth, markets can not self-subfficiently, and therefore in accordance with some notion of maximal efficiency, allocate capital, since capital must be applied by *organizations*, whereas market exchanges are 1) merely exchanges and 2) dependent on their relatively open, disorganized functioning. Organizations arise under market conditions 1) in order to organize production processes that are not simply provided for by markets, (but rather provide for the stuff of market exhanges), and 2) on the basis of the exclusion/capture of markets. (In the idealized neo-classical world of atomic producers and exchanges under perfect competition, there would be no profit above the rate of interest/cost of capital and thus no firms.) The illusion of the "autonomous" or automatic functioning of markets, resulting in some maximal efficiency, is the result of a logical inversion. Markets are precisely a mode of aggregating activities and decisions, which functions by reducing errors, their size, scope and duration, rather than by optimalizing some uniform standard of efficiency. As such, it will always be possible to realize profits on financial markets, provided one can successfully identify such errors. But those errors need to occur in the first place; that is, they are relative to the real organization of production. (Hence there are information asymmetries all over the place.) But the claim that financial markets over the long-run will out-perform any other method of allocation strikes me as simply a tautological reversion to the law of averages. There is a Darwinian/"devil's chaplain" aspect here: a good deal of capital gets wasted, in order for capital to function "efficiently".

As for James S.'s contention that decision-making should be decentralized and aggregated within organizations, yes, but isn't that what capital tries to avoid? Capital seeks control over the production and value-realization process, to the extent that capital is substitutable for labor and within the marginal trade-off between cost-reduction and control. Whereas I would have no illusions that production workers would seek managerial control and responsibility,- (that is another form of labor),- insofar as production and other workers control the production and realization process, they are precisely discouraged from exercizing such control, and having any input into the improvement of the process. This abuts upon questions of human and organizational "capital" and embedded systems-learning, but that is precisely what the demands of financial markets, (and the ambitions of top managers tied to those markets), tend to disrupt.

Just to add to the chorus, the real, (i.e. serious), debate is not between laissez-faire capitalism and state-directed centralized planning, but over a mixed economy, its proportions and composition and the modes and means of governmental structuring, regulation and intervention. I myself might be a bit of an outlier in such debates, but I will add that various political economies have historically emerged and evolved in such a way that they tend to pre-structure possibilities and preferences in that regard.

Is it possible that there exists qualitatively different markets? For example, physics has a model of an "ideal gas", and some gases fit that model very well, some, not so well. Hydrogen and uranium hexafluoride may require somewhat different equations.

As there is more than one kind of gas, there may be more than one kind of market. Compare car repair and medical services. If your car is broken, you may investigate what different mechanics have to say and what do they promise, for what price, while collecting info about their reliability. By the way of contrast, I am now incapacitated by a minor accident, and I would not do any research while I was writhing in pain. I went to my doctor, and as long as what he says seems to make sense, I will follow it. And the doctor was assigned to me by his medical group which was chosen by my employer. There is no "market freedom" here, and I do not see how it could help in the first place.

Similarly, market in crude oil futures may be a good idea while a market in electricity futures may be moronic. One is very liquid, the other almost absurdly non-liquid --- you cannot subsitute power in Visalia, CA at noon of June 23 with power in Fresno at midnight of June 24. Thus you may get an absurd number of minute markets and each of them can be easily cornered (even if it is not possible to corner all of them).

Is it possible that there exists qualitatively different markets? For example, physics has a model of an "ideal gas", and some gases fit that model very well, some, not so well. Hydrogen and uranium hexafluoride may require somewhat different equations.

As there is more than one kind of gas, there may be more than one kind of market. Compare car repair and medical services. If your car is broken, you may investigate what different mechanics have to say and what do they promise, for what price, while collecting info about their reliability. By the way of contrast, I am now incapacitated by a minor accident, and I would not do any research while I was writhing in pain. I went to my doctor, and as long as what he says seems to make sense, I will follow it. And the doctor was assigned to me by his medical group which was chosen by my employer. There is no "market freedom" here, and I do not see how it could help in the first place.

Similarly, market in crude oil futures may be a good idea while a market in electricity futures may be moronic. One is very liquid, the other almost absurdly non-liquid --- you cannot subsitute power in Visalia, CA at noon of June 23 with power in Fresno at midnight of June 24. Thus you may get an absurd number of minute markets and each of them can be easily cornered (even if it is not possible to corner all of them).

Yes indeed it could. The branch of economics that deals with this sort of thing is called "market microstructure theory", and was my own field, back in the days when I could claim to have a "field" without inviting derisive laughter.

D^2, would the (explosive) growth of hedge funds in the recent half-a-decade affect your critique of weaknesses of EMH? (As hedge funds try to seek out and exploit mispricings in the market?) Wouldn't part of the failure of EMH in the cases you quote be the thinness or lack of development of developing-world equity markets (or that the equity markets in the SE Asian countries were small in comparison to the developed world equity markets).

d, almost anything squared SHOULD have a field!

2559 You can buy viagra from this site :http://www.ed.greatnow.com

3893 Why is Texas holdem so darn popular all the sudden?

http://www.texas-holdem.greatnow.com

3217 Get your online poker fix at http://www.onlinepoker-dot.com

2165 black jack is hot hot hot! get your blackjack at http://www.blackjack-dot.com

Can anyone please define "market inefficiency"? And then based upon this definition, give me an example of "market inefficiencies" that existed in the past?