« The value of unfunded pension liabilities, and interest rates | Main | Don't bubbles burst when pricked? »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

"There is only one asset, land, and one consumer good, corn. No other goods, not even loans, or money."

Totally unrealistic.

How about this title "Bubble Warning; Why Assets are overvalued in the present (or near present) because of near 0% currency denominated debt from the few (mostly the rich or the investment banks)"?

Totally unrealistic

I'm pretty sure it wasn't meant to be realistic. The question is what complications you could introduce that would reverse the conclusion. If you can't think of any - that is, if the only role of those complications is to make things more complicated - then there's no loss in forgetting about them.

I interpret all assets are overvalued in the dynamic sense as a dearth of new worthwhile investment opportunities, that the economy is too static and not generating sufficient economic progress. One where both consumption is satiated and investment to produce more of the same is satiated leading to a lack of growth. One can say it is doing all that it can do, or one can say all that it is doing is insufficient.

I think perhaps that the Economist meant that financial assets (and some commodities used as financial assets) are overvalued. My understanding is that many in the financial community hold the view that these assets are overvalued relative to real assets i.e. labor and factories and general productive capacity/prospects.

Cash is an asset and the unit of account. Therefore no.

But I recently (unsuccessfully) tried to explain to Scott: bubbles form because those assets are considered a better store of future value than money. People switch to using money only as the medium-of-exchange and rely on the bubbled asset to act as a store of value.

When society starts coalescing around an alternative money (e.g., housing, tech stocks) the value of that asset increases due to its rising fungibility and liquidity. i.e., there is rational positive feedback.

But when something breaks that perception, the asset begins to lose its liquidity/fungibility which in-turns breaks that perception further. Until the asset reverts to its prior non-money-substitute status.

The answer is yes.
Asset prices represent wealth or past production, and labour income represents current production
If it suddenly takes 100 years of income to buy a house, the house it overvalued. It does not matter if it also takes 10 years of labour income to purchase a car. Asset prices are just a means of relating to current labour. It may take a hell of a lot of time if labour income is supplemented by net credit growth in excess of the declines real labour income which obscures the price signal, but they are interdependent.

Seems to me that subjectivity has an objective component, in that it is determined by objective neurological processes. To the extent that we can predict the outcome of these processes, we can make valid economic statements about subjectively priced assets' being overvalued. If you have good reason to believe that future art collectors will like Picasso less than current art collectors do, and that art speculators haven't taken this likely change into account, then you are making an economically meaningful statement when you say that Picasso's paintings are overvalued.

Similarly, if you think there are cycles in subjective time preference, and that people's preference for future over current consumption is likely to decline, and that investors have not taken this likely change into account, then you are making an economically meaningful statement when you say that all assets are overvalued, even if you think that the interest rate is subjective.

We are observing interest rates that abnormally low and idle capacity that persists in spite of those abnormally low interest rates. In some sense, if you think that interest rates will eventually return to more historically normal levels and that the idle capacity will eventually be used (or disappear), then, in some sense, you do think that all assets are overvalued. Personally, I wonder. For a sufficiently long time horizon, I agree that, in this sense, all assets are (probably) overvalued. Probably. But the question is the time horizon. Is it 2 years, 5 years, 10 years, 20 years, 50 years? Judging by Japan's experience (which seems to have expanded to include the rest of the developed world, rather than being resolved as one might have expected), I imagine that the time frame could be quite long. All assets can stay overvalued, perhaps, for much, much longer than anyone who speculates against them can stay solvent.

Nick,

Yes, the rate of interest is subjective. Once we start talking of all assets in the economy (and hence include credit risk) the subjectivity of the rate of interest is not contigent simply on the time preference of consumption but also on greed and fear. In fact, it is arguable that greed and fear feed-back into the time preference of consumption and affect that risk-free rate too.

Saying that all assets are overvalued simply means - there is too much greed. The stochastic discount factor is not high enough.

It might be more meaningful to discuss overvaluation of all assets classes on average rather than each individual asset. Clearly the perception of future value has an income and capital growth element.

For asset classes that are sought primarily because they return an income the question of overvaluation is likely to be obscured at present by the vastly different loan costs facing different market players - governments are providing capital at virtually zero cost to banks while individuals face significantly higher interest rates than was the case two years ago in most instances.

For asset classes where the primary investment goal is capital appreciation (art for sure, houses in Anglo Saxonia and Spain?) then the relative amount of leverage available may be more relevant than the absolute funding cost - and given that these investments are based on an expectation of future valuation by others would seem a priori to be more susceptible to bubbles as there is less of a concrete value anchor available for current analysis.

One element that may drive a secular shift in the perception of an appropriate rate of return is the increase in longevity - if I have longer to enjoy my Picasso/house, and will be enjoying returns form my investments/labours for longer I might be prepared to spend more on it/accept a lower annual rate of return on it.

Related program. Last night's Fifth Estate on CBC (Canadian equivalent to 60 Minutes, often picked up by PBS's Frontline) had an interesting program on the financial "magic" of Investment banker Lehman Bros. ultimately leading to its collapse due to the American real estate bubble - one it aggressivley created with others.

You can view online here:

http://www.cbc.ca/fifth/2009-2010/house_of_cards/

Too much Fed: Stephen provides a very clear explanation of why economists sometimes use "unrealistic" models.

Lord: In my simple example, there were of course zero new investment opportunities (you can't make land). Certainly the existence of investment opportunities (and the rate of return they would offer) would in the long run affect the supply of assets, and therefore the price. But given the existing supply, could we ever justify saying that *all* assets are overvalued?

csissoko: maybe that's what the Economist meant (I confess I was more intrigued by posing my question than interpreting the Economist). But houses and farmland are real assets too. And Canadian house prices are now back at about their previous peak. I have been trying to get data on farmland prices around the world; what I have seen suggests farmland prices are also very high, compared to historical levels. Stock prices have recovered a lot from their March lows, but not to their previous peak. It's only the safest and most liquid financial assets (like government bonds) that are very highly valued, compared to the past. I don't think anyone could claim that financial assets are generally overvalued compared to real assets, compared to history.

Jon: cash can be overvalued, even though it's the unit of account. It just means every other asset or good has too low a price, relative to money. Which means money has too high a price, relative to everything else.

And other assets almost always are a better store of value than money, on average. Money (at least currency) pays no nominal interest, and normally depreciates in real terms.

name: "If it suddenly takes 100 years of income to buy a house, the house it overvalued. It does not matter if it also takes 10 years of labour income to purchase a car." Why? what's magic about 100 years?

Andy: OK, there you have an argument. In my land/corn example, people's valuation of land would depend on their rate of time preference, but also on their expectations about future people's rates of time preference. And we could, in principle, say that their expectations were wrong, even if we couldn't say that their preferences were wrong (de gustibus etc.). But that would mean you would have to ask people what their expectations were (or figure out some way of distinguishing preferences from beliefs). You couldn't just look at land prices and say they are too high.

I think I disagree with you on the time horizon though. If asset prices do indeed stay high for a long time, that seems to validate, in retrospect, the expectation that preferences would not change.

On a related but slightly off-topic point: how do you know something is a bubble? Answer: if you prick it and it bursts, it was probably a bubble. If you prick it and it goes back to its original size, it probably wasn't. The experience of farmland, and Canadian house prices, is suggesting the latter. But maybe I'm pushing the "bubble" metaphor too far, though I'm tempted to do a post on this, if I can get my head clearer.

Ritwick: Normally I associate the word "greed" with an inability to delay gratification. "I want to consume now, and to hell with the future!". That would mean that greed would cause *low* asset prices. The two "subjective factors that determine asset prices are: time preference (patience vs impatience); and expectations (optimism vs pessimism). So if people's expectations can be shown to be overly optimistic, then yes, I expect we could say that all assets are overvalued, as in Andy Harless' argument. But "greed" is a pejorative word, that adds nothing to the analysis, and only confuses the issue. It's like saying "asset prices are too high because people are bad!"

ralaughton: I think you might be right that an increase in longevity in an overlapping generations model would have equivalent effects to a lower rate of time preference. Yes, if everyone had a constant probability of death d in each year, and a rate of time preference proper r, then I think the effective discount rate in an economy with no bequest motives would be 1/(d+r). But empirically, would the small rise in longevity be enough to outweigh the effects of demographics (a decreasing percentage of young people in the population)?

just visiting: thanks for the link.

I think that this is an example of the fallacy of composition: taking a simplified example and applying it to the complexity of the real world. In reality, there are 10,000 types of land and 10,000 uses of land. Calculating the real rate of interest from all the permutations of that 10,000x10,000 is *in theory* possible. But the calculation is so complicated, and would require making so many contributory assumptions along the way, that the time to do it would be close to infinite -- and the result would have changed by the time you got it. So lenders make guesses that are more or less informed by conventional wisdom -- except when wisdom has failed as in the past few years. Then they stop lending and say that all assets are overvalued.

The rate of interest is really an infinite number of rates based upon an estimate of future ability to pay/yield in an infinite number of circumstances. Overnight lending rates to banks differ from credit card rates to debtors with poor credit rating. It can't be less than zero, and they are all figured using some historical (though arbitrary) base, and working up or down depending upon estimates of future profit.

Not to be too picky, but even the simple example of land and corn ignores future variations in weather and the possibility of soil exhaustion -- or what the price of corn will be next year, or at least what the exchange value of corn will be relative to other crops. (In the real world interest is paid not in "corn" but "money.")

You can only assume a "real" or "objective" rate of interest retrospectively -- after the future has happened. Given the past couple of years, you can't blame anyone for questioning the current value of all assets, since no one knows what future profits from them will be. But the only answer to the question "are all assets overvalued" is "we'll see."


Nick you misunderstood my point is that cash cannot be overvalued at the same time everything else is.

Nick said: "how do you know something is a bubble? Answer: if you prick it and it bursts, it was probably a bubble. If you prick it and it goes back to its original size, it probably wasn't"

Nick, this is an elegant insight that can lead to more refined discussions on ‘bubbles’. Pls. do make your post on this as soon as you can clarify it a bit further. As everyone on this forum probably senses, everything now looks like a bubble, but is anything really a bubble? Pls. specify on your ‘pricking’ idea, like : what ‘bubbles’ should be candidates for pricking. ie, if we do not know until after pricking whether it is a bubble, how do we choose what to prick?; who to prick bubbles ie., should it be the government, central bank, private sector; when to prick a bubble ie., when it first seems to look like a bubble, or when an asset class has already risen too steeply beyond its historical norm; and how to prick bubbles ie., how do you prick it in a way that satisfies the bubble test (without causing unintended effects)

Assets produce income. Suppose that all (really most) asset classes can be temporarily overvalued relative to their expected ability to generate future income. That proposition requires strong assumptions about future willingness to pay for streams of income and the bounded rationality of agents.

Well-informed, rational agents should be able to see through the fog of distortionary monetary policy; I can't imagine how economy-wide asset bubbles would ever form in a world of perfect foresight.

P.S. Thanks for reminding us of the Cambridge Capital Controversy. Was it really as pointless as I seem to recall Solow suggesting?

Nick,

With regard to the changing demographic structure and its effect on the 'clearing' interest rate, if there are more older people (typically net lenders) and fewer youngsters (typically net borrowers) that would seem to provide an additional driver (greater supplier, lower demand) for lower interest rates.

Given the existing supply is what is at question. Given what we know, or at least take into account, can it ever be? It is what we don't know or take into account that changes things. The longer low rates persist, the less overvalued it seems. The longer the time horizon, the sooner population peaks and declines, the sooner wealth weighted population peaks and declines, the more growth declines, the more assets look under rather than over valued even if one expects them to lose value there is no other place to store it. This doesn't bar new discoveries creating new growth from appearing, drawing on capital, raising interest rates, and lowering asset values, but we have little to no means of anticipating them or taking them into account. I would not call what we cannot know a bubble, only what we knew and failed to reasonably account for one.

jcb: I would have argued that it's those who say "all assets are overvalued" who are making the fallacy of composition. Any one asset can be overvalued, if by that we mean overvalued relative to other assets. But if that's what we mean, we can't have all assets overvalued, any more than all kids can be above average.

I take your point that all assets have different risk profiles, and so won't all have the same rate of return in equilibrium. But comparing the different risk profiles of different assets is something we can usefully argue about. If I say land is riskier than houses, I can be right or wrong. But on time preference, there's no arguing.

Jon: Aha! Sorry, yes I did misunderstand you, and think I understand you now. If someone says "the money price of all assets is too high", then by definition the price of money cannot be too high. Yep. I misunderstood you because I was thinking that it meant "too high relative to consumption goods".

Rogue: Thanks! I had just finished writing a draft of a post on that point, and was wondering whether to post it or not, when I read your encouraging comment. So I finished it, and posted it, even though I'm not altogether satisfied with it. So, if it gets trashed I will blame you for encouraging me ;-) But I couldn't answer all the questions you posed, much as I would like to be able to.

westslope: I'm not an expert on the CCCControversy. There was an awful lot of fog generated. I eventually figured it out to my satisfaction, but I think a lot of people didn't. Cambridge US needed to add an upward-sloping supply curve of newly-produced capital goods. Cambridge UK needed to add an upward-sloping supply curve of savings. The simple Irving Fisher diagram really says it all about interest rate determination, if you can mentally switch it into multiple dimensions.

ralaughton: yes on the demographics, plus lower returns to investment if there are fewer young workers. That's part of Paul Krugman's Japan story, IIRC.

Lord: but you speak as if the rate of interest were something independent of asset prices. I would say that the rate of interest *is* the very same thing as (a weighted average of) the reciprocal of asset prices divided by their annual rental rates.

Nick, I don't think you have thought through your example:

The stock of land is fixed, therefore the marginal anything of capital is zero, as it is impossible to deploy an additional unit of land for any purpose. There no aggregate savings or investment in this economy. Both the stock of capital and total output are constant. Land here is just a proxy for saving at the individual level, but the only return available for saving is deflation (as money is withheld from the purchase of corn and prices adjust).

If you want to talk about interest rates, you need *productive* capital and money -- i.e. the economy as a whole must be able to produce more future output as a result of present investment, so that there is a return on that investment. The rate of interest will then be equal to this return.

Specifically, in your example the price of land will be such that there is no arbitrage possible between saving $X of money for 1 period and buying $X of land, holding onto it for 1 period, and then selling the land + crop at the end of the period. In equilibrium it cannot be possible to obtain free profits by choosing one option over the other. The only equilibrium solutions here are exponentially falling prices, or having land be of infinite value (so that no one will sell it).

While that is likely in the long term, it tends to vary with asset class in the short term. It certainly failed to be true of US housing in the last bubble. Stocks are much more volatile than interest rates which in turn are usually much more volatile than real property. Interest rates vary over the business cycle, with monetary policy, with inflation expectations. Whether a change is cyclical or secular can be difficult to estimate, and cannot not always be anticipated.

"I wasted too much time reading weird stuff, when I should have been getting my thesis finished."

This sounds familiar :)

Stephen Gordon said: "Totally unrealistic

I'm pretty sure it wasn't meant to be realistic. The question is what complications you could introduce that would reverse the conclusion. If you can't think of any - that is, if the only role of those complications is to make things more complicated - then there's no loss in forgetting about them."

I want to include all the currency denominated debt in the world that does not seem to be price inflating tradable goods and most wage income.

csissoko said: "I think perhaps that the Economist meant that financial assets (and some commodities used as financial assets) are overvalued. My understanding is that many in the financial community hold the view that these assets are overvalued relative to real assets i.e. labor and factories and general productive capacity/prospects."

I agree. IMO, they are overvalued because of currency denominated debt due to interest rates being too low.

RSJ and Lord: There is a conceptual mistake both of you are making. It's a common conceptual mistake, lots of economists make it. I used to make it myself.

We commonly say "low interest rates *cause* high asset prices". This is wrong. Low interest rates *are* high asset prices. The relation between interest rates and asset prices is mathematical, not empirical. We do not, and cannot, observe interest rates independently of asset prices.

An example: Take a 12 month Tbill that pays $100 12 months from today, and has a price $P today. *By definition*, the 12 month Tbill interest rate is the value of r that solves this equation: $P=$100/(1+r). And the interest rate on perpetuities is by definition the r that solves $P=annual coupon/r. It's exactly the same for all interest rates, except the formula is more complicated.

So the only interest rate that exists in my economy where land is the only asset is *by definition* r=(1 ton of corn/price of land measured in corn). (Land is a perpetuity which pays a coupon of 1 ton of corn per year.)

We only get away with saying "low interest rates *cause* high asset prices" because there are lots of assets, and so lots of interest rates, and so the relationship between the interest rate (price) of one asset and the interest rate (price) of a second asset is empirical -- an equilibrium condition dependent on substitutibility, rather than mathematical. But when we are talking about *all* asset prices, and *all* interest rates, we can't strictly do that.

Gotta prepare for class. I will return.

There is one problem with trying to equate them, asset prices are in the here and now, while interest rates are future expectations. If you are absolutely sure that land will always yield 1 ton of corn, that there will be no change in the future, that we have accounted for all possibilities, then there is no problem with equating them. Even in the presence of uncertainty we try to equate them, but uncertainty, change, our failure to account, can prevent that equality from holding even as we try to discount for them.

Nick,

Your response @ 6:27 wasn't directed to me, was it? I was merely pointing out that the derivative of a constant is zero -- not a mistake, in my book. Change your model to include variable levels of capital, in which the amount of capital deployed responds to investor return expectations, and you will have something meaningful to say about interest rates.

name:

The answer is yes. Asset prices represent wealth or past production, and labour income represents current production If it suddenly takes 100 years of income to buy a house, the house it overvalued. It does not matter if it also takes 10 years of labour income to purchase a car. Asset prices are just a means of relating to current labour. It may take a hell of a lot of time if labour income is supplemented by net credit growth in excess of the declines real labour income which obscures the price signal, but they are interdependent.

If all the value of all assets are somehow overvalued, then the value of labour becomes similarly overvalued (since labour creates/manipulates assets), and as such everything is pretty much equally affordable.

Plus there's what Nick says about 100 years being your magic number. There's no objective reason why 100 years rather than any other number, so it is a subjective thing as Nick's post said.

Someone who says that all assets are overvalued may as well argue that everything from the Industrial Revolution onwards has been one giant bubble for all the good it'll do them.

Whoops:

"everything is pretty much equally affordable"

should be:

"everything is pretty much equally as affordable as before"

RSJ: we may be misunderstanding each other. Maybe this will help. Introduce one-period real bonds (payable in corn) into my model. If land is priced at P tons of corn, and if that price is expected to stay constant (as it would, if time-preference stayed constant over time), then in equilibrium the interest rate on those one-period bonds would have to be 1/P.

Now, there is no MPK in my model, because there is no K. But there is a Marginal Product of Land; it's 1 ton of corn. And if we were able to produce new land, then land would become Kapital, the the MPK would itself become 1 ton of corn. My guess is that you may be confusing the MPK with the rate of interest. This confusion is precisely what was wrong with the Cambridge US approach. The MPK only determines r if the consumption good is identical to the Kapital good, so that C can be transformed into K at a relative price of 1. But if the marginal rate of transformation of C into K is not 1 (or not any constant), so that the supply-curve of newly-produced Kapital goods is an upward-sloping function of the relative price of K and C, then you cannot determine r from MPK. That was the bit that Cambridge UK got right. In other words, the relative price of Kapital assets to consumption goods in the Cambridge US approach was bolted down by technology. But that is a very special assumption, equivalent to a one-good model. And if that assumption were true, we would never see (real) asset prices varying relative to the CPI. But we certainly do see them vary. That's what this post is all about.

Lord: agreed. Under uncertainty, different assets may yield different rates of return/interest. Certainly ex post, but also ex ante, if people are risk-averse. But that uncertainty applies equally, in principle, to the rate of return on government bonds as to the rate of return on land. The rate of return on land is just as much *a* rate of interest as is the rate of return on government bonds, which is what we normally think of as *the* rate of interest.

Josh: seriously, do as I say, not as I did ;-). Old tenured guys like me can afford to take the risk of blogging the remains of our careers away!

"Now, there is no MPK in my model, because there is no K. But there is a Marginal Product of Land;"

"If land is priced at P tons of corn, and if that price is expected to stay constant (as it would, if time-preference stayed constant over time), then in equilibrium the interest rate on those one-period bonds would have to be 1/P."

Nope. Prices will not stay constant if there is a time preference other than 1. And Let b be the time preference. Then, then the price of land will be bP/(1-b), and land will depreciate in terms of money (not corn) at a rate of b. Corn will also depreciate (in terms of money) at a rate of b. Therefore the nominal return is zero for holding land for 1 unit. In the limit as b --> 1, then land is infinitely valued, and cannot be bought or sold. The rate of return on your "corn" bond is exactly the rate of return of land -- zero in terms of money.


No, K = Land here. There is no MPL unless you can change the stock of land somehow.

"And if we were able to produce new land, then land would become Kapital, the the MPK would itself become 1 ton of corn."

Now you are getting closer to the issue here. If we were to produce new land, there would be a *cost* in doing so. So I ask you to tell me how much labor and capital is required to add to the capital stock. If the answer is zero, and the production of land is costless, then the price of land is zero, right? In general, even though capital cannot be transmuted into consumption, nevertheless you need to pull real resources away from the production of consumption to the production of capital, and so you still have a hit on consumption in order to obtain more consumption in the future. That is how Solow defined the rate of interest, btw. But in order to talk about the rate of interest (or return), you need to describe how much capital and labor is needed in order to produce a unit of new capital (together with the depreciation rate), so that you can calculate the opportunity cost of producing more capital, and once you do this, you the rate of interest falls out, as in a competitive model, the cost of capital will be the return on capital. But you have not defined the cost of capital here, because it is impossible to add to the capital stock. Your model is exactly isomorphic to corn falling out heaven like Manna -- it is a model without capital.

RSJ:

Let me start with your last point first. Suppose 1 acre of land can be produced with 10 tons of corn, under constant returns. Then the rate of interest will indeed be bolted down to 10% (real), regardless of preferences. And that is indeed exactly the Solow perspective. And that is exactly the Cambridge (US) perspective, that Cambridge UK was correctly criticising. It is essentially a one good model, where corn and land are the same good, because one can be transformed into the other at a constant marginal rate of transformation (10:1 in this case). If you assume, for example, that there are diminishing returns to transforming corn into land, then the marginal rate of transformation depends on the level of investment, which in turn depends on the price of land, which in turn depends on time preference. This is the cristicism Joan Robinson (et al) made of Robert Solow (et al). (Except the Joan Robinson forget about the rate of time preference, so concluded the rate of interest was indeterminate).

Second point: "There is no MPL unless you can change the stock of land somehow."

Let's talk about the MP of labour instead. We can draw a MP of labour curve, and define MP of labour for each possible level of employment. Now, suppose the (aggregate) labour supply curve were vertical. The MP of labour is still defined. Exactly the same with land, which has a vertical supply curve. Exactly the same with Kapital, at a point in time, where the stock of Kapital cannot be changed by changing the flow of investment (until next period, of course).

First point. You sort of lost me here. But my bonds were real return bonds, paying a real rate of interest defined in corn. I didn't introduce money, or money prices. And the rate of time preference defines a marginal rate of substitution between today's and next year's *real* consumption (of corn, not money).

Nick: Let me start with your last point first. Suppose 1 acre of land can be produced with 10 tons of corn, under constant returns. Then the rate of interest will indeed be bolted down to 10% (real), regardless of preferences.

This is only true if time preference required period returns of 10% or lower, right? If time preference required 11% returns, wouldn't you have a real rate of interest of 11%, 9 tons of corn for every acre, and no new corn transformed into land?

dlr: yes, you are right. You would get a corner solution with zero investment, because the supply-price of new capital goods (10 tons of corn) is above the demand-price of land (9 tons of corn). The supply curve of newly produced land is vertical at 0 new acres, then turns sharply horizontal at 10 tons of corn, and the demand curve is always below it.

Unless of course the "10 tones of corn produces 1 acre of land" technology is reversible, so you can eat the land (which is what I was implicitly assuming). Then you would get an interest rate of 10%, and negative investment.

"If you assume, for example, that there are diminishing returns to transforming corn into land, then the marginal rate of transformation depends on the level of investment, which in turn depends on the price of land, which in turn depends on time preference."

Yes, all these prices depend on each other, and the net result of this is that price capital is determined by it's marginal product, which will "tend" to be constant. If you don't believe this, then you need to assume people are willing to throw money away by overpaying for the less productive land, and underpaying for the more productive land. *This* is what the Cambridge (UK) people were correctly criticizing -- the productivity of capital is not "bolted down" into a micro-style diminishing returns model. Larger companies have lower P/E multiples than growth companies, and people pay more for more productive land and less for less productive land, so that the returns from all forms of capital tends to equalize. You need to assume enormous stupidity on the part of investors to believe in a solow-style model that does not have constant returns to scale. One in which people not only incorrectly discount capital, but do so consistently over time so that future returns always surprise to the downside and those who purchased capital earlier (when there was less of it) are always guaranteed free profits simply from the growth in the size of the capital stock.

I agree with you that that all these prices are interdependent, but this does not mean that they are arbitrary. Over the long run, the return on capital needs to be the growth rate of the economy, otherwise P/E multiples will tend towards 0 or infinity. If the growth rate of the economy oscillates around a constant trend, then the return on capital will oscillate about this same trend value, as this is the only equilbrium value that allows all the relative prices to adjust in a way that does not allow arbitrage over time.

RSJ: There is SO much in what you say above I totally disagree with; I don't know where to start. Maybe we are totally misunderstanding each other, and talking at cross-purposes? It would probably be more productive if I do another post sometime, laying out Capital Theory as I see it.

Let me just take one example: "Over the long run, the return on capital needs to be the growth rate of the economy, otherwise P/E multiples will tend towards 0 or infinity. If the growth rate of the economy oscillates around a constant trend, then the return on capital will oscillate about this same trend value, as this is the only equilbrium value that allows all the relative prices to adjust in a way that does not allow arbitrage over time."

Suppose there is zero growth in the economy, and nothing ever changes from one period to the next. If people are impatient (if the rate of time preference proper is positive) then the rate of interest will be positive too, and P/E multiples will be finite. The rate of interest (assuming no uncertainty, and infinitely-lived agents) will equal the rate of time preference proper, and (assuming no depreciation) the P/E ratio will equal 1/r. (Exactly as in my land economy, where nothing does change over time).

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad