Demographics, the savings ratio and interest rates

It has been frequently stated that as populations get older (which is certainly happening amid rising life expectancy and low fertility rates), the savings ratio would fall and in turn (as there are less savings to buy financial assets) interest rates would be rising. As a result, longer-dated government bond yields are too low and will rise significantly – even irrespective of inflation developments – over the next years. The rise in the US savings ratio this year would then be only down to the fiscal stimulus and revert again soon.

This is down to the so-called life-cycle hypothesis of saving. Generally it is assumed that people below 35 of age do not save much (essentially because they cant), the next age group (35-50) is seen as having a moderate savings ratio and the pre-retirement age group is thought to have the highest savings ratio as they prepare for retirement. During retirement (i.e. roughly 64+ years), the savings ratio is seen to become negative as income drops. In turn, the savings rate in an economy should be the larger, the higher the ratio of people aged 35-64 relative to those >64 and <35 st="on">US, the so-called baby-boomers (which are said to be those born between 1946-1960) have been in the age group which saves the most but are just now starting to enter retirement age. With that the US economy’s savings ratio will fall over the next years and in turn, interest rates have to rise.

The development of the Japanese savings ratio which fell significantly during the 1990s is taken as prove for this theory as Japan has the largest share of old-age inhabitants.

I have several reservations about this kind of analysis:

1. if that kind of analysis were to be true, why should only interest rates rise (i.e. demand for bonds fall)? Essentially, in such an environment the demand for any financial/real asset should fall and take its expected return higher. So, if anything it would be an argument against any asset.

2. Age structure is not the only determinant of an economy’s savings ratio. For example, this paper The Japanese Savings Rate between 1960-2000 suggests that the development of total factor productivity (TPF) is far more important than any other variable. If TFP falls, then the savings ratio falls (and vice versa). To me this makes a lot of sense. A higher TFP means that there are more profitable investment projects around. In turn, companies will increase their investments and in order to do so increase demand for saving funds via higher interest rates. As a result, households can achieve higher returns by saving than in a low TFP environment and c.p. should save more. However, while a falling savings ratio amid an older population would see yields rise, a fall in TFP depresses the savings ratio and lowers interest rates.

3. The provision of social security and the retirement age are also important factors. One reason why for example, Germany’s savings rate did not fall earlier this decade can be seen in the changes to social security where amongst others, the retirement age is slowly rising while the formula used to calculate pension entitlements has become less favorable. A higher retirement age means that a) employees will stay longer in the high-savings-ratio state and less generous state pensions means that private pensions become more important. Given that the state of public budgets is very challenging almost across the globe we should expect that more countries will increase (or at least try to increase) retirement age over the longer term.

4. As I stated already in Collapse in Private Pension Funds’ Assets a key reason for a secular rise in the savings ratio?, the people near retirement are those hit the hardest by the financial crisis. We should expect that especially people near retirement (and parts of the retirees) which have relied on private financial sources for their retirement are forced to save more. This can happen either via spending less or via working more (and retiring later).

Overall, I am not convinced that the US savings ratio is headed lower and with that interest rates higher over the next years already due to the baby-boomer generation entering retirment. First, their asset base has been hit hard and more baby-boomers will be forced to work longer and delay retirement. Second, the fiscal deficit situation suggests that over the longer term, retirement ages will have to be increased in a growing number of countries. Third, the development of total factor productivity seems to be a more important factor driving the savings ratio but with opposite effects on bond yields.

Overall, I do not see demographics standing in the way of a long period of historically low government bond yields amid muted real growth and subdued inflation pressures.