A theory of return-seeking firms – Part 1

In neoclassical market theory firms are profit-maximisers.  That sounds intuitive enough – until you realise exactly how economists define profits and the problems it creates for the whole market model.

In this series of posts I will develop my ideas about an improved market model incorporating return-seeking firms at its core, rather than profit-maximising firms, can overcome some of the key shortcomings of the neoclassical market model.   All comments and criticisms are welcomed.

The first shortcoming is the shape of the supply curve.  The neoclassical model assumes they slope upwards both for a firm and for the market as a whole. Yet at best they are flat, and most likely they are downward sloping over any useful period of analysis.

A simple explanation is that market supply curves are the horizontal summation of supply curves for each firm.  Unfortunately this means that supply curves for each firm must be bizarrely steep in markets with very high numbers of firms if the market supply curve is still to be assumed to be upward sloping. Indeed, at the limit of an infinite number of firms, the individual supply curve must be vertical to maintain an upward-sloping market supply curve, and individual firm production would be zero.

Economists get to assume away problems like this by invoking the idea of a ‘short run’, where capital is fixed and maybe these assumptions could hold, versus a ‘long run’ where the supply curve is flat and so is the demand curve, and you basically have no model left.

So how does the central concept of profit-maximisation lead to such confusion? First we need to consider what economists mean by profits.  Basically profits are revenues minus costs.  But remember, there is no time in the theory. Imagine a discounted cashflow model with time periods out to infinity, then discount all rows of future period cashflows back to present values.  At what rate do we discount? The normal rate. Where does the normal rate come from? Essentially no one knows. But that is a question for a later Part in this series.

Leaving the issue of time to one side for a moment, we can then work out using this single period the revenue minus the costs, and call that profit.  So far so good. This is exactly a Net Present Value (NPV) calculation at the moment.

But do firms really maximise NPV? This is a serious question that lies at the crux of the whole theory of markets.

Consider a firm investment decision.  Option 1: Spend $100 to generate revenues of $110, making a $10 profit.  Option 2: Spend $200 to generate $211 revenue and $11 profit.

The profit, or NPV of Option 2 (since these are already discounted values) is $11 compared to $10 in Option 1.  If an economist was to advise this firm they would say take Option 2 because you make more money. But of course twice the funds are required initially to make that extra dollar.

Option 1 however, has a higher internal rate of return by construction.

To be clear, the NPV value is the time-condensed version of revenue minus cost, measured in dollars.  The IRR is the revenue minus cost per cost per time, or more simply, the percentage return per period.

I fail to see how investors can be return-seeking[1], squeezing the most out of every dollar they invest, while firms somehow behave with a different financial objective.  Surely if investors buying the firm are return-seeking then firms themselves would operate with that objective.

The economic rejoinder is that since there are no other positive profit investment options in equilibrium, firms are better off spending the extra $100 to make $1, than nothing at all.  The only alternative option of no returns arises from the assumption that all capital is in already in use elsewhere. So you must squeeze all you have out of your current capital.  Unfortunately this usually requires more labour, but the model also says that labour is fully employed elsewhere.  In fact, there is no curve as such in the neoclassical model, just the point where production takes place.

But in the real world capital is not fixed. Investing in new capital[2] is a decision regularly made by all firms.  And as Sraffa so clearly described, the output of many markets is the capital input to another – the production of commodities by means of commodities. Thus capital cannot be fixed except in the most extreme short-term conditions [3]. And under these conditions, the short run neoclassical model might be of use.  But firms won’t generally plan to operate in this way, and this only occurs during unforeseen short term “shocks” where prices are able to be adjusted quickly.  This is a rare situation that is nested within the broader model of return-seeking firms.

To take the neoclassical model seriously you also have to believe that we are in, and always in, equilibrium to make that case.  Even if one market is out of equilibrium, the neoclassical model has nothing to say about profits in any other markets – either all markets are in equilibrium, or none.

But enough about the shortcomings of the neoclassical theory of markets for the moment. Just what does this mean for the next wave of economic theory?

It means that the level of output at which a firm produces is far less than predicted by the optimal decision to equate marginal revenue with marginal cost.  As Alan Blinder’s famous research showed, 89% of firms operate on the downward section of their cost curve, despite the theoretical prediction that this would only occur in natural monopolies.  But in a model of return-seeking firms, producing on the downward-sloping section of cost curve maximises returns if firms face a downwards sloping demand curve for their product.

Compared to the neoclassical model, real firms seem to over-invest in capital and/or under produce.

The two graphs below show this principle using the fundamental cost curves for a firm with a given level of capital [4].  When a firm follows the economist’s advice they choose a lower rate of return on their investments but higher output.

Yet I know one situation where the economic model holds – property development.  Because the property developer has the monopoly on their land, maximising NPV also maximises their return since they can capture all the surplus.  If they have the option to spend $100 to make an extra $1 above a ‘normal profit’ they will.  This situation actually complies with the assumptions of the neoclassical model; that capital is fixed, in this case land, and firms will maximise the profit associated with that capital. This situation is just one of many that nests within the framework of return-seeking firms.  But the distinction becomes very important when I get to optimal timing of irreversible investment choices (when to build) for land owners in a future post.

Unfortunately it is difficult to translate even this minor change into the broader neoclassical theory without also revisiting and modifying (improving) many other critical assumptions in that theory. In future posts I hope to consider a few other critical points.

1. Where do profits come from? In the neoclassical model the normal rate of return is assumed to arise from some other market. But fundamentally that other market will be modelled with the same assumption of normal returns – the circularity problem means there is no origin of profits in the economic vernacular.

2. What are economic rents versus profits versus returns, and why don’t rent arise in the neoclassical model.

3. Competition is assumed to have already occurred in the model, and it offers little insight into the competitive opportunities available to firms. Competition is merely a static choice of output level.

4. The fundamental model has no important implications for asset markets and their impact on new physical capital investment.

fn[1]: I use the term return-seeking because maximisation usually implies a single unique solution.  However the realities of uncertainty and the complex interactions in the economy mean that at best firms “seek out” high returns.

fn[2]: I mean capital in terms of inputs into production that don’t physically end up in the product, such as buildings, machines, vehicles, tools etc, as opposed to material inputs

fn[3]: Depending on the market.  Some industries with large scale capital projects might temporarily see these conditions (such as resources).

fn[4]: Horizontal axis is output, vertical axis in first graph is cost/price in $.  The second graph has two vertical axes – return in %, and profit in $.  The scale of the two lines are not comparable, only the positions of the peaks.

Please share this article.  Tips, suggestions, comments and requests to [email protected] + follow me on Twitter @rumplestatskin


  1. I’ll bite. Have you read “Why Most Things Fail” by Paul Ormerod?

    “Game theory” is probably a better way to analyse the behaviour of the actors in markets, than equilibrium theory.

    • Yep. I’m a big fan of Ormerod, and read most of his work now (I went through a bit of an ‘Ormerod phase’ for a while). I’ve also replicated some of his network models myself.

      I guess I’m going for a unifying principle of markets, that better captures real patterns of production than the disjointed family of neoclassical models. The principle of return-seeking firms is really one that leads to agent-based style of aggregation, which may or may not involve network formation. Simple linear aggregation is absolutely not suitable for a model of markets. But that’s a topic for another day.

      Yep, I also agree about game theory as a useful tool. However you have to be careful about the huge difference between one-shot and repeated games. In repeated games almost any cooperative coordination is possible, ala the Folk Theorem.

      • I think I’ll just “go Hayek” and say that these things are “too complex” to try and formularise.

  2. “…..you must squeeze all you have out of your current capital….”

    Have you considered the effect of corporate taxes on this? Government can take the money at the point it is distributed as income to shareholders anyway; why take any at all before the point at which the company has even decided to distribute it or plough it back?

    Where else does real economic growth come from if not from ploughed-back profits? Even borrowed money for expansion, has to be paid back out of profits. Company tax is therefore a tax on economic growth. There is no rate that is “too low”.

    This is why the lower it is, the better the effect on the economy of the country that lowered it.

    Imagine too, how many businesses that fail in their first few years, might have survived had they had the profits taken off them by government, available to “plough back” instead as their business grew. Company taxes squeeze the ability of growing new-starts to finance stock and debtors and so on.

    I have not seen a convincing argument yet, that the government will not get the revenue anyway in the long term as profits are distributed as income. Any system needs ways to minimise avoidance of tax, but there is no point penalising everyone by setting up the entire system to close a loophole that a minority of companies might have abused. Better just close the loophole by targeting the method of avoidance.

    Why subsidise tertiary education “because it is good for the economy”, and tax ploughed-back company profits? Why not give tax credits for ploughed-back profits?

    • “Company tax is therefore a tax on economic growth. There is no rate that is “too low”.

      We might be agreeing a lot today Phil, but on this I only partially agree.

      A lot of what is accounting profits for companies is also simply economic rent. Look at the ASX – it is essentially owners of capital extracting rents – miners, banks, property owners etc. very few companies actually operate in the dynamic innovative sphere of real competition. Thus corporate profits taxes unfortunately combine the good parts of rent taxes, with the bad parts of profits taxes on those innovative companies.

      You would agree that eliminating profits taxes in favour of direct taxes on rents would be an excellent solution (and simplify labour taxes as well).

      • Ah yes, absolutely. I am clear about this; I meant the companies that do actually “produce” and add to wealth, rather than “seek” it via zero sum rents and negative sum rents.

        +100 for your lucid definition. You got my point exactly, and added the essential discrimination that I left out. Taxes on economic rents instead of profits is a superb idea.

    • So you suggest that all costs of the required infrastructure for companies to operate successfully be externalised onto the citizens with no taxes (or only those on “rents”) on companies.

      Roads, railways, the legal system, the army, police, … all of which contribute to the abilities of companies to make a profit ought be paid for by companies as well.

      The shareholders of companies benefiting from infrastructure are not all citizens, many are foreigners. You are suggesting that Australian citizens and permanent residents ought pay the full price for the physical and social infrastructure so taht the companies of foreigners can accumulate more wealth, while the additional taxes then needed to be paid by the citizens and residents will ensure that they accumulate no wealth.

      There would also be a massive transfer of wealth from the citizens/residents to those who own shares including foreigners. If companies paid no tax their value would increase from say 16 (PE ratio) times 60 (after tax earnings after tax of 40) to 16 times 100 if there was no tax of 40. If the discounted cash flow model of company valuation is correct then the value of the after tax earning capacity of citizens/residents must diminish by the same amount assuming no reduction in services (which is a separate debate).

      I think your proposal is totally misguided in an open economy that welcomes foreign investment.

      • Explorer, you miss my point that the government gets the tax revenue anyway by taxing distributed profits. But it does matter if they take this money off a company before it even distributes it.

        No company is ever going to “not pay a dividend” just so it and it’s shareholders can avoid paying any tax at all – what would be the point of buying shares in such a company? If any company did act in this way, its growth would be such that the wide range of other taxes paid by it and its staff and customers and suppliers, would make the foregoing of “company” tax by the government worthwhile anyway.

        Of course I do not think that shareholders offshore should not pay taxes on profits distributed to them. We don’t need a destructive tax per se instead of a better designed system.

        • No company is ever going to “not pay a dividend” just so it and it’s shareholders can avoid paying any tax at all

          Ohh boy, that’s just not true at all. You even referred above about game theory

          A proprietry company with all shareholders being employees and under 35 will most defiantely find a way to not pay tax and still consume personally.

          • They’ll have to pay income tax on their “salaries” and perks won’t they?

            Heck, the whole tax system relies on honesty and/or auditing. There is no excuse for a major harm-doing tax “just because the alternative is exploitable”. All taxes are “exploitable” if it was not for the chances of being found out and penalised.

            Even under the status quo, there is really no difference in the situation you describe. The incentive to try and make personal consumption a business expense is just as great.

          • A director with a wage of one dollar, won’t pay much income tax. Retained profit can be held, then find itself into the share price, which execs awarding themselves shares will only be ‘paid’ in capital gains, not income.

            A ‘justifiable’ company car with be paid for with pre-tax dollars.

            Every dinner can be ‘entertainment’.

            I don’t disagree with reducing company tax, I just find there has to be more of a level playing field with it, and income tax. The are both taxes on exertion and enterprise.

      • Besides, land taxes would be the best solution of all to getting the beneficiaries of infrastructure spending actually paying for it. The system is already full of distortions that are gamed by rent-seekers.

        “…. costs of the required infrastructure for companies to operate successfully….” should in any case be borne by the land owners that the companies rent from, not the companies themselves. The infrastructure is not attached to the company, and does not move with it if it changes its location; it is attached to the site.

        There will be exceptions to this, such as the need for unique infrastructure for a particular industry; but even then the tax can be applied to the land rather than the businesses located on it. Then it does not matter what changes there are in business structure and how profits are shifted around to try and avoid the appearance of benefit at one particular location.

    • One of the problems with reducing company tax is that a lot of the profits do not get taxed as income in the hands of individuals, because they are sent to overseas owners.

      • PS, I now see you have addressed this point above. A withholding tax on dividends to foreigners would achieve the same result.

  3. “…..Yet I know one situation where the economic model holds – property development. Because the property developer has the monopoly on their land, maximising NPV also maximises their return since they can capture all the surplus. If they have the option to spend $100 to make an extra $1 above a ‘normal profit’ they will. This situation actually complies with the assumptions of the neoclassical model; that capital is fixed, in this case land, and firms will maximise the profit associated with that capital. This situation is just one of many that nests within the framework of return-seeking firms. But the distinction becomes very important when I get to optimal timing of irreversible investment choices (when to build) for land owners in a future post….”

    I hope you make the distinction between urban economies with planned growth boundaries (or proxies for them) and those without.

    In the latter case, the investor who owns land can decide to withhold it from the market, as its value will increase as the city grows around and beyond his land holding. I call this “positive sum rent seeking”.

    In the former case, the classical economic assumption that the market allocates the fixed supply of land to its best use, is revoked. The “fixed supply of land” under the classical economic assumption is VAST. In the case of Australia, it is “all of Australia”; and on top of even that, there is the law of import substitutability to consider. As long as a nation can freely import food, it can have insufficient farmland to feed itself, yet the farmland can remain priced at international prices for farmland, and hence eminently convertable to urban land by the operating of the free market.

    But the supply of land within an urban growth boundary (or a proxy for it) is NOT “vast” at all, and is easily “cornered” by even one single actor in the urban economy unless other actors join in a bidding war to stay in business. This is compounded by the fact that land owners then have far greater incentive to withhold their land from the market than the previously described “positive sum rent seeking”. This is simply because the land is rising in value orders of magnitude more rapidly than waiting for the city to expand.

    I call this “negative sum rent seeking”.

    The land that is placed off-limits to the urban economy by the UGB (or proxy for it) would have been easily accessed by the urban economy due to the low cost of travel. The cost of automobile travel is low enough that the amount of land economically accessible by an urban economy is many more decades worth of “supply” for growth than it is worth anyone’s while “cornering” and withholding from supply.

    • Considering basic urban economics: the land value curve for an urban economy slopes up from fringe to centre based, among other things, on the cost of transport saved by a more central location.

      Where there is no UGB and no bidding war for the quota’d supply of land, this curve’s slope is gentle and continuous. It actually extends miles out into the country beyond the existing fringe, and the amount of land involved is massive; like 100 years or more “supply” for urban growth. If some speculator “corners” all the land within the first mile or 2, which is quite some feat, then competitors can simply “leapfrog” the holding. It is only 2 minutes drive on an open road anyway, and about another $1,000 worth of real estate value versus “transport costs saved”.

      This is how it works in cities without an Urban Growth Boundary.“Affordable housing” is basically an issue of whether or not people can “leapfrog” land bankers holdings to build themselves houses. When there is no UGB, as in the affordable US cities, no-one bothers to land bank. The “supply” economically accessible to the urban economy via a few minutes and a couple of dollars worth of further car travel, is far too big for anyone to “corner”.