People respond to rising asset prices through so-called wealth effects.
Small movements in asset prices may have little effect, but if the rise in wealth is large enough then, after a while, some people change their behavior. If the stock market has soared, perhaps they cancel their reg- ular savings plan and use the money to go out to dinner more often, boring their friends with their skill in picking stocks. Others may take some profits and use the proceeds to buy a new car or a boat. If house prices rise fast they may increase their mortgage to spend money on a new kitchen or a home extension.
Some might say that people are foolish and shortsighted if they immediately spend gains. But many people have a target level of wealth and, if asset price inflation enables them to reach it earlier than they expected, why not spend more? After all, for most people the purpose Chart 1.2
The NASDAQ bubble
Source: DATASTREAM
of acquiring assets is for spending at some point. Of course, if the increase in prices is temporary and later reverses, they will be in for a rude awakening. There is also a danger that, after a period of price gains, they start to expect continuing gains at the same pace and adjust their spending further upward.
The evidence suggests that most people do not immediately spend gains but in fact respond only gradually. Possibly they are slow to real- ize that they are better off. Or perhaps they take a cautious approach to higher asset prices and only spend the gains when they believe that they are permanent. There is a potential trap there, though. The judg- ment as to whether or not higher asset prices are permanent tends to be based more on whether prices hold up for a while rather than whether valuations make sense. But as a bubble inflates, prices often exceed sensible valuations for an extended period and people start to see those high levels as normal.
Another common response to higher asset prices is to increase bor- rowing to finance higher spending. In the US it is relatively easy for peo- ple to borrow against stocks, even with only modest stock portfolios. In other countries often only those with a large portfolio can directly bor- row against stocks, though there are other ways to take leveraged posi- tions, including CFDs (contracts for differences), ETFs (exchange traded funds), and, in the UK, spread betting.
Still, for the average person, borrowing against assets is far more likely to be in the form of “mortgage equity withdrawal” (or MEW), as house price gains are released by remortgaging. Over the last 20 years it has become much easier to do this in many countries as rules have been relaxed, either through further advances or through switching mortgages and increasing the amount.
If the money raised from borrowing against assets is used to fund spending, then the effect is a fall in the household savings rate and it is this change in the savings rate that is the measure of the wealth effect.
The savings rate is calculated as the difference between current income and current spending, and therefore ignores the fact that the increased spending may only be possible through new borrowing or sales of assets.
A crucial point to note here is that a fall in the savings rate only has a one-off effect on the growthof spending. Imagine a couple who respond to
a rise in the value of their house by cutting their regular savings plan from 10 percent of their annual income to 5 percent. As a result there is a one-time 5 percent increase in their spending that year and a 5 per- cent drop in their savings rate. But the following year, if they stick to the 5 percent savings plan, their spending will only change in line with their income.
This creates a tricky problem for monetary policy, because central bankers are very much focused on the growth of the economy. If every- body in the economy cut their savings rate by 5 percent at the same time, there would be a sudden leap of 5 percent in consumer spending.
This would immediately set alarm bells ringing and create fears that the economy was growing too fast and that higher inflation would follow.
The response might well be to raise interest rates to try to calm things down. But the next year consumer spending growth would drop straight back to its old growth rate. If central bankers fail to realize what is happening, there is a real danger of a monetary policy mistake, with interest rates set too high.
The risk becomes even greater if people finance the extra spending from borrowing or from sales of assets, rather than from changing their regular savings, because in both these cases the extra spending will actually be reversed the following year. Suppose consumers raise enough cash through sales of assets to increase their spending one year by 5 percent. Next year, unless they repeat the exercise, they have only their income, so spending will register a fall of 5 percent. The central bankers are even more at risk of getting it wrong.
Sometimes people respond to rising wealth by taking on more debt to finance new asset purchases. One common reaction is to buy a sec- ond home, perhaps selling some stocks for a downpayment but also tak- ing out a new mortgage. In this case there is no change to consumer spending or the savings rate or to consumers’ overall wealth (after sub- tracting the new debt). But total assets are up, debt is higher, and the consumer has more risk. Meanwhile, home prices are likely to rise fur- ther, pushed up by the new demand.
Fortunately, some people are more cautious in response to higher asset prices. For example, a risk-averse response is to sell some stocks to pay off debts. Others may respond to rising house prices by
remortgaging with a larger loan, using that debt to pay off other (higher- interest) loans, or by increasing deposits, against a rainy day. Again, risk is reduced. If debt is paid off, as in the first case, then risk is certainly reduced. In the second case though, the individual’s overall wealth is still dependent on house prices holding up. And from the point of view of central bankers there is still the worry that, one day, those extra deposits will be spent.
All the effects described above go into reverse if asset prices fall enough to seriously reduce wealth, so that savings rise as a percentage of income. This last happened in the 1970s when the ravaging effects of inflation on real wealth encouraged a rise in the savings rate as peo- ple tried to restore their balance sheets. But also, when asset prices fall, people often get scared and want to reduce their risk, by paying off debt or switching out of riskier assets such as stocks. Of course, this is not the best moment to do so and indeed may be exacerbating the asset price cycle. Nevertheless, many people are impelled to reduce their risk.
CALCULATING WEALTH EFFECTS
Economists have tried to calculate the size of these wealth effects in practice. This is not easy, because rising asset prices usually coincide with rising incomes and falling unemployment, both of which also encourage spending. Moreover, to some extent rising asset prices and ris- ing spending may both be the result of a monetary stimulus from the central bank, via a fall in interest rates. Nevertheless, despite all the caveats, most of the research finds that there is a measurable effect on consumer spending from higher asset prices, but that it varies from country to country.5A few studies failed to find any wealth effect from stock markets and a very few found that house prices were not very sig- nificant either. However, most find house prices more important than stock prices, with the effect up to twice as great.
Research in this area goes back to the 1970s and in the US a rule of thumb has emerged saying that for every one dollar increase in asset val- ues, consumers will spend 5 cents more. Most subsequent studies for the US seem broadly to support this figure, though it is perhaps better
to think of a range of from 3–7 cents.6 The link between asset prices and consumer spending seems to be greater in the US than in conti- nental Europe. Americans are generally more influenced by stock prices since more than half of the population own stocks, either directly or through mutual funds, while an increasing number of employees have 401K pension accounts.7
Sometimes it is argued that stock market wealth effects may not be very important in the economy because stockholdings are concentrated among the top 10 percent or so of earners. The trouble with this argu- ment is that so is a large part of consumer spending. In Britain stocks probably have less direct impact, though there are substantial indirect holdings in pensions and insurance policies. But housing prices in Britain are very volatile during the economic cycle and this has histor- ically been very important, as we shall see later.
In continental Europe, stock markets are generally much smaller in relation to GDP, with companies relying more on banks for finance.
And in some countries, Germany and France for example, house prices have been rather less volatile than in the UK, often because the finan- cial system or tax structure discourages easy buying of houses. In coun- tries where the financial systems are liberalized, as in Scandinavia in the 1980s or the Netherlands and Spain in the 1990s, there have been substantial bubbles. Intriguingly, one study that found only very lim- ited stock market wealth effects did find them for the US, Ireland, and Finland. These are the three countries with the most dramatic stock market booms in the second half of the 1990s.8 This would appear to confirm the idea that wealth effects may not matter very much in nor- mal times, but can become very important during extreme movements.