Is the stock market wealth effect a “wealth illusion”?
In “Beyond Doughnuts” I argue that when interest rates are already low, further cuts may raise desired saving – the opposite effect to what most macro models assume. Something related had also been troubling me: the stock market wealth effect. Central bankers argue that one of the ways QE works is through higher stock prices producing a wealth effect. Seems obvious? Stock prices go up, you feel wealthier, you spend more …
But there’s a problem with this story. Stocks are not really that different to bonds (they just have uncertain cashflows), so when the price of stocks goes up, the yield goes down. There shouldn’t really be much of wealth effect. To illustrate this more clearly let’s think of someone retiring in one-year’s time who holds all their savings in a 10-year government bond. The coupon on government bonds is fixed – so they get paid the same dollar amount each year regardless of what the price does. So why would person change their behaviour because the price rises: they will still have the same income in retirement? Robert Merton makes a very similar point here.
The same is true of the stock market – when the price goes up, the expected return falls. For example:
Assume that everyone who owns US stocks is planning to retire in 20 years time. Let’s say the S&P500 is priced to deliver a 4% real return over the next twenty years. Now, 4% compound over 20 years is a total return of 120%. That is a reasonable basis for retirement planning.
As it happens, even though 4% is a very reasonable estimate of the average real annual return, it is virtually certain that the stock market will not return a steady 4% real each year for 20 years. For example, it might rise 100% in the next 12 months. Would that make stockholders richer? An individual can sell to a greater fool, but collectively this is zero sum. So if the stock market doubles in 12 months that will not help anyone retiring in 20 years – because the subsequent 19 years’ returns will be poor, and the cumulative 20-year return is unchanged.
So the purely ’rational’ view of the wealth effect has to be that some people are selling stocks to buy goods and services. Maybe there are those in retirement who sell stocks every year to pay for their consumption. In their case, they could respond to a fall in the expected return by spending more (it’s another version of the Euler equation). This is quite a risky strategy, though, and requires a belief that the gains will be enduring. Also, those they are selling to should be raising their savings rates.
As it happens, I think there is a stock market wealth effect – partly because people don’t fully realise that rising prices equal falling expected returns, but also because a fall in the equity risk premium implies that things are getting better. Collateral values influence lending conditions, and banks look at household balance sheets on a mark-to-market basis. The reality of cyclicality is that if people think things are getting better – they are. If you also believe that Say’s law should be reversed and supply is endogenous … things are even better.
Warren Buffett famously asserts that if you are a net saver you should celebrate falling asset prices, because you can accumulate assets at a higher expected return. Some smart investors practice this, but everyone feels better when they think they are winning.
A 4% real return expectation not unreasonable. The PE multiple of trend reported for the S&P500 is around 25x, equivalent to a 4% earnings yield – a decent proxy for the expected real return.