Updated: 2005-03-31
Created: 2004-05-08
In principle transactions are symmetrical; one party exchanges some valuable(s) with another party.
It is however common to prevaricate as to whether the
transaction is a sale or purchase by/for whom, by calling a
transaction a deal
, generically; in some cases,
usually involving barter or currency exchange, it may be genuinely
non obvious which party is the seller.
There is a simple rule to determine which party (or parties) are selling and which are buying: the party (or parties) receiving the more liquid valuable(s) are the seller(s).
It is important to note that the liquidity of the valuables exchanged in a transaction depends on the context, such as the place or the time.
For example, when changing dollars into yen, the party receiving the dollars is the buyer if the transaction happens in Japan, but the seller if the transaction happens in the USA. If the transaction happens in neither country, it all depends on how liquid each currency is in the country in which the transaction takes place, and in this particular case the dollar is usually more liquid than the yen.
For example, when exchanging labour for money, who is the seller depends on how tight the labour market is; if it is easy to hire labour, the seller is the company, if it is easy to find a job, the seller is the employee.
In general this distinction matters because the seller as a rule has to work harder than the buyer to close the transaction. Considering the principle above, the reason is obvious: as rule, being more liquid is preferable to being less liquid, and the buyer needs to be persuaded to accept a less liquid valuable in exchange for a more liquid one. Also, often sellers enter into the transaction to obtain a profit, and buyers to obtain some kind of use from the purchase.
I reckon that one of the most amusing details of corporate and public life is just how stupidly structured are the incentives given to agents by principals.
Among many, the most hilarious and common practices are that usually incentives are related only to upwards movement of the target quantity, and to the unnormalized level of that quantity.
Most incentives pay the agent a prize if some increase of a target quantity above some threshold is observed, and the prize is usually some percentage of the increase.
This gives agents a very large impulse to ensure that increases do happen half of the time, even if then bigger decreases happen the other half of the time. Rather than have stock prices or sales or another target quantity be static over four years, and collect nothing, it is far better to pull all the stops in the first two years, show increases of say 10% a year, and then suffer two years of 10% decreases, ending up at 98% of the initial level.
It is also very nice to be given prizes on target quantities not normalized for trends; this means that if trends are favourable, prizes will be earned without having to work for them.
Part of the agent-principal conflict is that information is asymmetrical, as the agent usually has lots more inside information than the principal, and that the ability to shape events is also asymmetrical, as the principal's almost sole power is to change the agent, while the agent can change the way the business is run, or even, and more easily, how the target quantities are calculated or even defined (with a little lobbying).
Ideally instead of goal-oriented prizes there should be a variable element of compensation; the variable part should be related to the absolute level of some quantity, not to positive increases only, and should become negative when the absolute level falls.
The quantity should be a normalized quantity, usually normalized by trend, both across the industry and across time, so as to remove from it both industry-wide trends, and temporal trends, which would punish the agent undeservedly if they turned negative, just as now they are used to provide undeserved rewards when positive.
Also, ideally the quantity should be directly related to cash, because notional values (like earnings, or orders booked) are too easy to manipulate for insiders.
In part because of the booming Chinese economy and its impact on raw materials markets there have been warnings about impending inflation; and in part because of the continuing weakness of the level of salaries driven by greater international competition in the labour markets there have been worries about deflation.
Also, in the first half of the 2000s decade, there have been waves of what should be called very high asset price inflation, first for stocks, then houses and eventually bonds, at the same time as the prices of manufactured goods and consumer services are steady or declining.
These observations are not contradictory, and can probably be to a large extent explained with the shift of a large percentage of GNP (probably because of political and legal actions) from labour to capital.
Also, in many countries the after tax situation seems to have improved mostly for high income, high saving taxpayers, and for those that derive income mostly from capital.
The shift in before and after tax income from low to high incomes has had two effects:
Therefore one would expect that prices of things purchased by low income consumers would fall, such as cheap manufactured goods, and the prices of things purchased by high income investors would rise, such as financial assets.
Thus one can have at the same time high inflation at the high end of incomes, and substantial deflation at the low end.
The overall issue is then not whether inflation or deflation are happening, but inflation or deflation of what.
Also, inflation or deflation have been traditionally defined in terms of absolute prices. But this can be regarded as simplistic; the core of a definition of inflation or deflation is related to the purchasing power of people's resources not merely of money.
People perceive inflation as a swing against them of the terms of trade. Those for which the main resource is their work will therefore look mostly at the difference between the rates of change of the prices of their work and of consumer goods, those for which the main resource is their money they will look at the difference between the rates of change of the prices of money and of other assets.
Also, interest rates are often considered a signal of inflation; very high absolute interest rates are regarded as being symptoms of inflation.
Perhaps, but there is the difficulty that absolute interest rates cannot be zero or negative; also high absolute interest rates may often have other causes, like a scarcity of financial capital.
It looks more fruitful to look at the level of real interest rates, as a negative real rate of interest accompanies inflation, being both a symptom and a cause of it. And again this depends on the inflation of what because the real interest rate depends on the difference between the nominal one and the rate of change of prices of something.
So the real interest rate can be at the same time positive in terms of cheap consumer goods and negative in terms of expensive assets, suggesting inflation only in the latter.
Looking at the price of IBM stock over the past 40+ years there is cause to feel very worried. The current price of IBM stock seems amazingly high considering it is twice what it was then IBM was the most profitable monopolist in the IT world, and now it is mostly just a consultancy business.
But then if one compares with the NASDAQ index it looks like the NASDAQ is still at incredbily high levels, and that the recovery in prices is a bit out of line with historical trends.
I reckon that the two important questions these graphs prompt are:
In addition, it looks like government bonds went out of fashion for USA investors, around 1995 probably as retirement accounts were switched massively into stocks, for various reasons, and non USA purchasers eventually replaced domestic investors when the USA switched to a war budget in 2002 and thus a policy of high deficits and easy money.
Also, Internet stocks have fallen, but not that low. If there is an Internet stock that is largely Cisco, and stock price is currently oscillating around the $20 level, which it had reached in 1999: and in 1999 it was one of the hottest momentum stocks around, and people who bought the stock at $20 in 1999 did do betting that it would continue to go up and up, and it did, reaching almost $70 in 2000. People who buy Cisco today are more or less betting that it will stay at $20. Strange times.
It has become popular for listed companies, especially technology companies, to provide returns to shareholders as capital appreciation via retained earnings, rather than the distribution of dividends.
This is largely motivated by taxation, as capital gains are almost universally taxed less than dividends. A smaller reason is that companies should retain earnings in case they are later needed to fund investment, as raising funds is more expensive than retaining them.
The practice of not distributing dividends has however two very large negative consequences:
The second point requires some explanation. If earnings are distributed as dividends, the value that a company has to an investor can be derived from the amount of cash it distributes, which is easily measured. The only major trick that can be used is the classic one of liquidating capital into earnings (for example by insufficient depreciation charges) and then into dividends.
But if earnings are retained, the value of a company to an investor becomes solely that at which its stock can be sold; and if all investors are in the same position, each of them is in the awkward position of having to guess what the others would buy his stock for, so company valuation becomes a circular guessing game about other investors' guesses, instead of estimating the expected stream of cash distributions.
In order to put a thin veneer of seeming rationality on the
guessing guesses
game, valuation has been
tied to various quantities, instead of the easily measured one
of cash distributed as dividends.
These quantities, most of them variously clever intepretations of the idea of earnings, are notional in the sense they are accounting categories which depend on how things are classified, and they don't have the clarity of cash.
This is very convenient for executives; it provides them with an extraordinary incentive to play with accounting categories and rules, as such manipulation directly affects the value of their stock options, and it is much easier than manipulating dividends, as these have to be distributed in cash, which requires either hard work to generate profits, or hard risks in running down the company's capital.
Even better, the game can be played until credulity ends, then clean the act, and then restart it, as shifts in classification don't actually alter the fundamentals by themelves.
Many investors, intent on timing the market on a momentum strategy, are happy with management playing with accounts: if they can get out before the other investors realize it's all been a game of words and numbers, they can make the same profits executives make on their options. However even as to this management have much better information as to how far they have gone.
Caeteris paribus
Other things being equal
is a premise often
used in arguments, especially in economics. It has the advantage
of allowing to focus on one aspect of the argument and its
specific contribution to the outcome.
Unfortunately, very often other things cannot be equal; one variable cannot be changed while keeping the others constant, because they are correlated, either directly or indirectly.
There are even famous theorem's, like in economics Kaldor's on
income distribution, that are based on an application in a case
where the premises of the theorem make it impossible for
Caeteris paribus
to happen.
The phrase should be rewritten as If and when all other
things are indeed uncorrelated and do remain unchanged
.
The most important work of Piero Sraffa has been to show that under some simplified but tolerable assumptions it is possible in way of principle to build a measure of value that does not depend on the distribution of income among producers, which is the crowning achievement of value theory.
An even less known result of the same work is that under many plausible conditions, the less efficient a producer is, the higher is its income.
This sounds highly counterintuitive, but it does happen, and happens because the price of a product depends on relative scarcity.
If producers of good A become more efficient, and producers of good B become less efficient, and demand remains the same, the price of good B will rise more than proportionally relative to the price of good A, because good B is now relatively and perhaps even absolutely scarcer.
This depends a bit on an assumption of other things being equal, which of course is not necessarily the same.
To give an example that I hope looks very realistic, consider the case where there are 100 lawyers, each of which completes 10 cases a year for a yearly income for each lawyer of 200,000 pounds, or 20,000 pounds per case.
This basically means that there is enough (and more than enough probably) demand from the public to have 1,000 cases a year prosecuted at a market clearing price of 20,000 pounds each.
Now suppose that the efficiency of the lawyers is halved, that is each is only able to complete 5 rather than 10 cases a year. What will happen to the income of each lawyer? Will it be halved to 100,000 or will it be the same at 200,000 or will it become higher?
Odds are that it will stay the same or probably become higher, that is the price of each case will more than double.
There was proven demand from the public to have 1,000 cases a year completed, and the value of each having each case worked upon by a lawyer was at least 20,000 pounds a year, because the public was willing to pay that much.
But now of the potential demand of 1,000 cases at 20,000 pounds each, only half of that demand will be satisfied; of the 1,000 cases that the public would have reckoned it was worth pursuing at a cost of 20,000 pounds each, only half will be able to be pursued, so the public will bid up the price of each case, until the market clears, that is until the price of each case will be high enough that it will be too expensive to pursue the least valuable 500 cases.
Given that the potential 1,000 cases there will be a wide variation in value, and the only thing we know is that they are all worth more than 20,000 pounds to pursue, and that Zipf's law usually applies, it can well happen that the 500 most valuable cases that will get pursued will have a value greater than 40,000 each, in which case the income of each lawyer will have risen in absolute terms, even if their efficiency has been halved.
The halving of the efficiency has resulted in each case becoming a much scarcer and thus potentially more valuable service, and thus lawyers who sell that service can command a much higher price per service, and in many or most cases, where the relative scarcity is particularly high, a more than proportionally higher service.
In the same way, if everybody else becomes more efficient, the income of lawyers will probably grow too. The reason is that if everybody else is more efficient, then the prices of all goods or services other than legal cases will become lower, and it can happen that therefore most people will have to spend less money as consumers on everything but legal cases, and thus they will have more money to bid up the price of legal cases.
In the cases where 100 cases are pursues annually at a cost of 20,000 each, that means that there are at least 1,000 cases worth more than 20,000 and there are enough people with at least 20,000 pounds available to spend on legal expenses.
But if the number of people with at least 20,000 to spend on a legal case grows, and each lawyer is still only able to complete 10 cases a year, and there are more than 1,000 cases worth more than 20,000, then the notional demand for legal cases will become greater than 1,000 and the price of each case will rise until only the most valuable 1,000 cases are actually completed each year, perhaps at a price of 30,000 each, if the value of the 1,000th least valuable case is 30,000 pounds.
Note that obviously the above discussion is highly simplified, and in particular there will be many intermediate cases. The most important assumptions are that the number of lawyers will not expand in reaction to demand, and that the efficiency of lawyers as a category will be the same. The above discussion applies only to lawyers as a whole, not to each one; if the average lawyer can complete 10 cases a year for 20,000 pounds each, an individual lawyer whose efficiency is half the average will have half the average income.
However, the general principle is that in the absence of increased competition, and with unsatified notional demand, decreases in the absolute or relative efficiency of a category of producers will result in their unit prices, and usually their absolute incomes, growing.
The use of lawyers as an example here is partly intentional, as in those countries which have huge unsatified demand for legal services and high barriers to entry, either formal or informal, to the legal profession, the prosperity of that profession depends on the level of inefficiency it manages to achieve, and usually they try pretty hard to lower their productivity as a category; pettifogging leads to higher profits.
But in general there are statistics that show that sectors of the economy, usually in the services, that are relatively sheltered from competition and for which there is significant unsatisfied notional demand, and where productivity has probably been stagnant, have seen their prices rise much faster than average inflation in the past few decades, as they have become relatively less efficient than other sectors where relative and absolute productivity has risen faster, and whose prices have risen much less than the average.
Such sectors are for example health care and university education (where, unlike in the UK, the market for university education is not rigged), where cost of treatment and tuition fees have been rising way ahead of average inflation as demand for care and education have risen significantly thanks to higher disposable income in the economy at large, and productity and absolute capacity have not risen at the same rate.
