The Brexit generation

The consequences of economic post-traumatic stress disorder and the associated desire for safety are equally important for countries such as ours, where macroeconomic quantities such as growth and inflation can be volatile for long intervals, and policy changes are often subject to prolonged uncertainty. In particular, the concept may provide insights to help address the important and longstanding problem of excessive gold holdings.

Policymakers have long been frustrated by the fact that Indian household savings are disproportionately directed towards unproductive assets such as gold, rather than channelled into formal financial markets. An investigation of the most recent All-India Debt and Investment Survey (AIDIS) (with Cristian Badarinza (NUS) and Vimal Balasubramaniam (Oxford)) from 2012 reveals that on average, households in India hold more gold than they do financial assets, and the median gold holding is roughly ten times the median holding of financial assets. In the international context, Indian gold holdings are exceptionally high when compared with countries such as the US, the UK, or Germany, and while Chinese households hold comparable fractions of non-financial assets to Indian households, their gold holdings are lower than those of Indian households.

Why should we care? For one, Indian households stand to reap numerous benefits from redeploying resources from gold to formal markets, including a more favourable risk-return profile, diversification benefits, and liquidity. On a more macro level, there are potentially big wins from a tilt away from gold towards more productive assets, such as a reduction in the current account deficit from declining gold imports, and the mobilisation of much-needed funds for uses such as infrastructure.

What explains Indian households’ gold holdings? AIDIS data show that as households become more educated, their allocation to gold remains relatively constant, but their allocation to financial assets increases steadily, meaning that gold is less important in a relative sense. Wealthier households also eschew gold, tending to hold real estate instead. Households with children, especially female children, tend to hold higher amounts of gold, and there is considerable state level variation — for example, households in South India hold greater amounts of gold. These factors clearly suggest that cultural factors are also an important part of the picture. Yet there is still much variation in gold holdings that is left unexplained by these determinants.

This is where the notion of post-traumatic stress disorder becomes useful. A frequent explanation for high allocations to gold is that Indian households use it as “inflation insurance,” seeking an alternative store of value to cash. But this conjecture is hard to prove, since the available history of aggregate data is too meagre to draw convincing inferences. One way to surmount this challenge is to use state-level variation in inflation combined with state-level variation in gold holdings. This approach allows a clearer view — the data reveals that households in states with historically high levels of inflation volatility tend to hold significantly higher amounts of gold. Digging deeper, it turns out that this connection is especially strong if the head of the household experienced high levels of inflation when they were young (age 25) — such households turn out to hold particularly high levels of gold.

This finding of “post-traumatic inflation disorder” is intriguing. Over and above a host of other factors, experiencing periods of high and volatile inflation when young is associated with households’ propensity to hoard gold. The effect of this experience is considerable, comparable in magnitude to the effect of having children.

Gold holdings appear to have deep roots in household fears and expectations arising from lived experiences of high and volatile inflation. The long-lasting nature of the impact on asset allocation considerably strengthens the case for a strong inflation target to tackle inflation volatility at its source, in addition to solutions already in place such as gold monetisation and gold bond schemes. A broader lesson to be drawn is that uncertainty per se can have undesirable effects many years into the future — meaning that the costs of delay in policy implementation may be significantly higher than previously thought.