St Andrew’s Study Trust lecture series: Living with Uncertainty; 4 September 2012

The notion of uncertainty – and risk, which as we shall see, is something else – is central to economics and the economies it describes. Of course there are external events which generate shocks like the Lisbon or Canterbury earthquakes. But today I am going to talk about the financial shocks which the economic system generates. It’s a complex story, so I am going to have to leave some important bits out to keep within the time.

Financial shocks occur almost exclusively in modern market economies: bubbles, panics, depressions and recessions and global financial crises are rare in traditional societies. One of the central characteristics of a market modern economy is that money is pervasive. Economists have long been aware that money, and some different but related phenomena which I shall call “financial paper”, are central to these shocks. Today’s presentation is about how they are.

While some talk of the “mystery of money”, the only mystery to an economist is why people get confused about it. Perhaps they do not think rigorously. For an economist “money” is the “medium of exchange”, what people are willing to accept in exchange when the sell something or settle a debt. There are other functions of money – a standard of value and a means of deferred payment – but these are convenient consequences of money being a medium of exchange.

I have not time to go into why some things are money and others are not. Perhaps in a modern economy the test is that if you can pay your taxes with something it is money. That includes coins, notes and cheques.

There are also things which are “near money” – that is they may be easily converted to money. If you have a bill – or taxes – to pay you can usually go into your bank and transfer from your savings account to your current account and write the cheque on that. A savings account is near money; a cheque account is money, you can pay bills directly with it.

That something as a dollar value attached to it does not mean it is money or near money. You can also sell your house to obtain the money you need to pay a bill. That is far more complicated, so it is hardly sensible to call your house “near money”. It is an “asset”. It is important for the story I am about to tell, that you are uncertain as to how much cash you will get when you sell your house. That is different from your bank deposit; you know exactly how much it is worth.

A particular group of assets are “financial paper”. They are contracts to pay a certain amount of money at some time in the future. They are neither money nor near money,

Perhaps the simplest example is an IOU, which is a contract – a promise to pay a certain amount in the future. If the IOU is from your much trusted brother-in-law who is a billionaire you are pretty certain it will be repaid. If it is in your balance sheet as an asset it is in the promisor’s balance sheet as a liability.

You may even sell it to someone – although its value might be discounted a little because the purchaser is not as trusting of your brother-in-law as you are. Much financial paper is not nearly so secure. For instance some New Zealanders have deposits in finance companies which have gone bankrupt. Whatever the face value of the deposit, the actual return will be zilch if the finance company is unable to pay it back. Sure, there is a contract that says the company will pay – give you money – in exchange for that deposit with them. But if they cant; the contract is worthless.

The value of an IOU is the money you can convert it into. That is different from an asset like a house. A house is a house is a house; it is something tangible with an intrinsic value. Certainly houses also have a monetary value, but an IOU, like all financial paper, has only monetary value.

IOUs and deposits in finance companies are simple forms of financial paper. There are many more complicated ones. In a modern financial system some of the most important financial paper are contracts where the pay-outs are dependent on contingent events.

An early form of contingent contracts was life insurance, which involve one party (say a person) paying another (a life insurance company) a certain amount, with the second party agreeing to pay a somewhat larger amount if a particular person died before a particular date. The reason they could offer such a deal is while we cannot be sure when a particular person will die, based on past experience the life companies have a good idea about how many in a pool of people will. So the risk of an individual dying can be calculated and the event insured against.

Is there a fundamental difference between investing in a contingent contract and gambling – between a contract insuring a life and a contract that if a thrown dice comes up even you get paid and if it comes up odd you pay? There is not a lot – the mathematics are much the same. The distinction is often made according to the usefulness of the insuring transaction. It would seems prudent for a firm to insure the life of its chief executive, since there would be great difficulties if he dies suddenly. But what about someone else betting on the chief executive dying? Maybe they are a shareholder in the firm, maybe they just want to gamble.

This becomes very pertinent when you can insure against a firm or a country going bust. There have been recent examples where the amounts insured far exceed what might be lost if the event occurs. Some must be gambling.

Insurance may be the simplest form of financial paper, that is a contract whose payment depends on a contingent event, but a whole raft of increasingly complex ones have developed; among them are “shares”, “futures”, “options”, “swaps”, “hedges” and “derivatives”. In each case some parties make a contract whose payout depends upon whether a particular event occurs.

We get mesmerised by the mind-boggling sums attached to the value of a particular piece of financial paper. However, behind the row of zeros is the simple idea that there is a contract in which one party has agreed to pay the other an amount if certain circumstances occur, and that one of the parties thinks that the value of that contract is this huge number.

It may not be, of course, since the contingent outcome may be different from what the valuing party assumes. But if enough of this financial paper is held in the investor’s portfolio, the wins and losses average to the average value of the portfolio.

Clearly it is critical how the financial paper – the contingent contracts – is valued. If this can be done in an objective way, so that we can all agree on its value, then the financial paper can be bought and sold. This liquidity makes the financial paper more attractive to investors, who may need the cash before the events in the contract have played out.

Over the years economists and others have developed very sophisticated techniques for valuing the contingent contracts – the financial paper – developing an analysis called “portfolio theory”. The valuation requires an assessment of the probability distribution which underpins the contingent events. I would lose many of the audience if I went through the mathematics. So let me just say it depends upon knowledge of a couple of parameters which underpin the probability of the various outcomes of the contingent events. (I’ll just squeeze in that they are the mean (or average) and the standard deviation.) Once given them, you can value a particular piece of financial paper and make an assessment of whether it should be in your portfolio.

But suppose you don’t know those parameters. John Maynard Keynes pointed out you then suffered uninsurable “uncertainty” rather than insurable risk. Risk is about known unknowns; uncertainty is about unknown unknowns

If there is uncertainty you cannot know those key parameters. Portfolio theory does not work. However some Bayesian statisticians said you could use your best guesses. The dispute is a bit complicated – you can read about it in Skildelsky’s Return of the Master.

But there are two key points. First, suppose everyone is wrong about the parameters they guess for the distribution. Then there may be major distress when the contingent event becomes a reality.

Second, and more subtly, the relevant parameters may be literally unknowable. I wont go through the details but the mathematician Benoit Mandelbrot observed ‘10 sigma’ storms in financial markets pointing out that “[a]ccording to portfolio theory, the probability of these large fluctuations would be a few millionths of a millionth of a millionth of a millionth….But in fact, one observes spikes on a regular basis – as often as every month – and their probability amounts to a few hundredths.” He warned of ten sigma storms a decade before the Global Financial Crisis, but no one listened. Why?

One kind of answer is that portfolio theory may have been imperfect but neither Keynes nor Mandlebrot offered an alternative so portfolio theory continued to be used. This is a nice illustration of the scientific truth that theories get replaced not because they are known to be wrong, but because some one finds a better theory.

Additionally some people became very rich from portfolio theory. They have a vested interest in defending the theory, and since we give great weight to the judgements of the rich on a whole range of issues outside any competence they have, everyone assumed that the portfolio theory that made them rich must be right.

Another kind of answer as to why we ignored the warnings was that portfolio theory seemed to work reasonably well up till then. Of course there had been the odd crisis; the dotcom bust, the Enron debacle, the failure of Long Term Capital Management, a firm whose board of directors includes a couple of economists who were Nobel laureates for their contributions to portfolio theory. Most of the time, the financial system manages to get through these crises. Firms are always falling over, contracts are always going wrong, most businesses have a list of bad debts. But the system had muddled through, and it was assumed that it would continue to do so.

Yet as Mandlebrot predicted, there would be occasions when there was perfect storm in which the financial system breaks down into a systemic crisis. The 2008 Global Financial Crisis was one. It was not the first systemic financial crisis – that is usually said to be the Dutch Tulip Bubble in the seventeenth century; there have been many more since. I doubt that the GFC will be the last. It is only unusual because of its size, reflecting the increasing global inter-dependence of the financial system. When the tulip market crashed in Amsterdam there was barely any impact on London just over the way.

There is perhaps an even more fundamental reason why financial paper has become so important. Karl Marx shrewdly summarised the changes in the evolving capitalist economy in the nineteenth century by a change in the role of money. Once. he said, the economy could be represented by

C –> M –> C’.

That is you started of with a commodity C (perhaps you made it yourself) which you converted it into money (M) in order to convert into another commodity C’(which you would rather have). Money in this circuit was a means to an end, not the end itself.

However Marx observed that as the economy evolved, a new circuit developed:

M–> C –> M+,

Here one starts with money, converts it into a commodity which in turn is used to convert into more money. That reverses the roles. Money becomes the end and commodities are a means to that end. Monetary values begin to supplant intrinsic values in the way we think about the economy.

More recently the commodity in the circuit has been replaced with financial paper.

M–> FP –> M+.

You no longer have to produce commodities to make money, you can do it by producing financial paper – contracts about contingent events.

How can that happen? The conventional wisdom said it arose because the financial system was organised around allocating risk between investors – there is no Keynes or Mandelbrot critique of the existence of uncertainty in the conventional wisdom. It said that some low-risk investors were willing to pay others – high-risk investors – to take over their more risky contracts, just as you will pay a life insurance office to reduce the risk of your partner living in penury if you die early.

I have not heard that explanation used since the 2008 Global Financial Crisis for it seems a bit naive when a lot of low-risk investors were hammered while some high-risk investors were barely touched. But even before then, high-risk investors and those transacting on their behalf seemed to be making a lot more income than could be justified by the marginal differences in risk they seemed to be trading.

So let me offer an alternative explanation as to why financial paper in the money to money circuit seem to generate profits in the good times? Financial paper are contracts about events in the future. We may attach probabilities to the outcomes of those events but until they actually happen we dont know the true value of the financial paper. In principle the combined positive and negative values of the financial paper ought to be zero in the world’s balance sheet. When someone has an investment portfolio with a billions dollars of financial paper as assets, somewhere else there are other balance sheets with the same financial paper as liabilities. In your household balance sheet there may be a deposit in a bank of $10,000. In the bank’s balance sheet there is an offsetting liability of that $10,000. When we add the two together they exactly cancel each other out.

But because financial paper involves guesses about the future they may be valued in different ways (with different parametres) the net position in the world’s balance sheet may be increased by financial paper. We have to guess at the underlying probabilities; different people make different guesses and the combination of their guesses may not obey the rules of probabilities. The mismatch can add to the sum of the world’s wealth in the short run.

This is such a startling phenomenon that I need to explain it step by step. I use an example of a housing bubble. Suppose one day someone authoritative announces that the value of all houses in New Zealand had doubled. All house owners dutifully enter the new value in their balance sheet – if they were meticulous enough to keep one – and feel they were much richer. Are they? This is only a change in the monetary value of your house; what it is worth if you sell it. The doubling of the market value of your house has little effect on its intrinsic value. You may say “I am twice as rich”; but you are unlikely to say “I am enjoying living in the house twice as much”.

A complication is mortgage debt, which acts as a brake on rising house prices. If the brake is weak, as is it was in much of the last decade when banks were flush with overseas funds, house prices rise as home owners try to show off their wealth with more grandiose homes, and as they try to increase their wealth from the capital gains which come from leveraged purchases in rising market.

One of the fundamental rules of economics is Stein’s law, which says “If it cant last, it wont.” House prices cannot rise forever, relative to ordinary prices, so at some time after a boom they will stagnate or even fall. You’d have thought investors could figure that out, but research increasingly shows that their behaviour is not entirely rational. Again this is a huge topic in its own right, so I shall have to skip it. If you are interested, read psychologist Daniel Kahneman’s Thinking Fast; Thinking Slow, and economist Robert Shiller’s Irrational Exuberance.

It is extraordinary just how irrational investment decisions can be. It probably arises because if one make an unwise investment decision one rarely get an immediate signal that it is stupid. Without early feedback, learning is dampened and, in any case, just about everyone else is doing it, so how can it be foolish? Or so it seems until the day of the crash.

The same applies to financial paper generally. Now you might think that while ordinary humans are unsound investors, surely those in the investment industry are not. It turns out they are just as unsound us – but richer.

Should not their judgements be disciplined by these powerful mathematical models they use? That only applies if the models are correct – recall that Keynes said they are fundamentally flawed because they pretend uncertainty can be converted into risk – and if they models have the right assumptions – Mandelbrot says they dont.

But surely the financial paper cannot add to net wealth since the contract values should net out? Only in the long run; in the interim the two sides of the contract may value them differently. A simple example may be that the NINJA – no income, no job, no assets – who borrows to purchase a house with no intention of paying the debt back and values his side of the mortgage in the balance sheet as zero. But the lender is likely to put the mortgage into her or his book at full – or perhaps partial – value. So the values do not net out when the two balance sheets are combined.

Unlike a house, there is no intrinsic value in the financial paper. Yoga Berra might have said the intrinsic values of financial paper is not worth the paper they are written on. Their value is the money you can turn it into by selling it to someone else. So nobody finds out the real value of the contract until it is settled. The US housing finance market started looking soggy in 2006 when the NINJAs began failing to pay.

Because there is no way to resolve the actual value of the financial paper – the contingent contract – until the settlement happens in the future, it is possible for the net value of the financial paper in the balance book of the world to be positive rather than zero. Investors act on the apparent, but deceptive, increase in wealth, which leads to bullish market activity we call a “bubble”. At some point reality undermines their speculations based on false assumptions, and the bubble pops, the financial system collapses and the economy goes into a downswing. During the reversal the false profits have to unwind, which is why we expect a long recession after a major crash. (We are already in the fifth year of the recession associated with the Global Financial Crisis, much longer than the shallow speculators thought in 2008; some informed commentators think it may last another five or more years.)

I could extend this story through to the details of how the Global Financial Crisis occurred, but instead I want to make four analytic points.

The first is that people become exposed to the financial paper without understanding how sound – or unsound – the underlying contracts are. It happens all the time. You dont know much about the bad debts that the bank you deposit with. You believe they are under control and that the equity of the shareholders provides a cushion to protect you. That may be broadly true for New Zealand’s conservative banks, but that was not always true for overseas instances.

The second point is that dealers profit as the financial paper was bought and sold. In the main their profits are offset by the losses of those on the wrong side of the contract. However the latter happens some time later so the dealers seem to be making a profit out of thin air; what they are really doing is taking over some investors’ savings.

Third, the system would work a lot better during the downturn if people had not borrowed to fund their investments. The problem in a housing boom is not the values of the houses no matter how ludicrous they are, but that people have mortgages which they used to pay the fictitious values.

The final point is that some of that debt was taken on in actions which were more like gambling than investment. As I said the distinction is not clear. Indeed the vernacular describes betting on the TAB as an “investment”. Now I have no problems with someone who punts on horses for recreational purposes, provided they do not bet more than they can afford. They may pretend they will win, but generally they don’t – the punting is for the fun, giving you an incentive to follow the horses more closely. The problem arises when you bet your house on a horse, or take out a mortgage to bet on some piece of dark horse of financial paper.

I am not the first to draw attention to the gambling in the financial system. When it was a lot less complex than it is today, Keynes waspishly described the share market as a casino, remarking “When the capital development of a country becomes a by-product of the activities of a casino, the job is likely to be ill-done.” However the problem is not the gambling per se, but the gambling with other people’s assets, because that is what debt is – other people’s assets.

Should the gambling occur? Let me begin with an analogy from the Canterbury earthquake. It was a consequence of the Indo-Australian and Pacific tectonic plates ramming into one another; earthquakes largely occur near plate boundaries from releasing some of the titanic forces as they clash. If those two plates were not crashing into one another, the Zealandia continent would be almost entirely under the sea. It was heading that way 22 odd million years ago, until the Pacific tectonic plate began expanding. The smash into the Indo-Australian plate drove up part of the sunken Zealandia into the land mass we call New Zealand. If that had not happened we would not be here; all there would be is sea and there would be no earthquakes. They are the collateral damage of living in a New Zealand.

Financial crises are kinds of earthquakes. They too are a consequence of an upheaval – the market-money economy. No markets, no money, no financial crises. But we would be much poorer for it.

We may think that Africans are poor, and of course they are. But even their material standard of living is today about three times higher than it was two hundred years ago.(Ours is about twenty times.) That was the consequence of the market economy which enabled specialisation and encouraged innovation. But its downside is the occasional financial crisis.

Such crises are an inherent feature of our economic system, even if we restrict people from gambling with other people’s savings. The failure may well be compounded by stupidity of investors, by what economists tactfully describe as “misalignment of incentives” which sometimes amount to fraud and corruption, and by poor economic management. But the crises will happen even if these failures are addressed – perhaps less often, perhaps less violently, but they will occur.

Certainly we should address them as far as we are able with the humble recognition that human ingenuity tends to outwit the most careful institutional designer (and our designs are not always careful).

Perhaps we should think about to whom we should give the most protection. I am old-fashioned enough to think we should give first priority to children, the sick and the old – as Michael Joseph Savage once articulated. He would not be satisfied that we achieve his ambition today.

Where I might differ from Savage, is the realisation that the government has not an unlimited ability to protect everybody, to insulate the entire economy from uncertainty, except at a horrendous cost to our material standard of living. For as long as there is economic uncertainty there will be financial crises. That may be a cruel reality, but what a boring world it would be if everything was certain.