How are salary growth and equity returns related? In the short term, they're not. The correlation between yearly increases in UK average earnings and equity returns is practically zero.
This doesn't rule out a longer-term connection, though. Consider a drunk man walking his dog on a lead. He staggers left and right, while his dog darts from side to side. In the short term the movements are uncorrelated. But in the long term the movements are correlated because the dog's lead means he can only be so far from his owner. Mathematicians call this cointegration.
Could equity returns and salary growth be cointegrated? We, and some academics,1 think so. Moreover, a link between salary growth and equity returns is plausible as labour and capital investment are linked by economic growth.
Interesting stuff, but how is this relevant to DC pensions?
In a DC context, the ‘human capital' of an investor is the present value of future contributions made to DC pots. Contributions are typically a percentage of salary.
Human capital can be a huge asset to investors, as it represents a flow of future cashflow from their jobs. It complements their financial capital - the money currently invested in pension pots. Typically, young investors have little to no financial capital but bucket-loads of human capital. By retirement the situation is reversed and they rely on financial capital for their pension.
A common assumption is that contributions that underpin human capital are inflation-linked. That makes modelling relatively easy and makes some sense - we should expect salary increases to be related to inflation. But if this is all we model, it treats human capital as 100% bond-like, akin to a UK index-linked gilt.
Why might this matter? Well, how bond- or equity-like human capital is has implications for how to invest financial capital. The more bond-like human capital is, the more aggressive investors should be with their financial capital and vice versa, in our view.
We've built a model similar in concept to the academic papers sourced below, but tailored to the UK.
As expected, it shows salaries are more linked to equity returns the further into the future one goes. If you're within five years of retirement, your future salary and contributions only stretch out for five more years, and how they change bears little relation to equity markets. In that case, your human capital is pretty bond-like.
But if you are far from retirement, many of the contributions you may make in the future are linked to equity returns. That means human capital, which is the present value of those future contributions, is more equity-like.
Implications for investment strategy
Under some reasonable assumptions we believe the ideal percentage of total (i.e. human plus financial) capital invested in equities should be constant over time. The chart below shows how extreme the impact of moving to a model that recognises how equity-like human capital is depends on how close the investor is to retirement:
In this example, there is a complete reversal of strategy when the investor is young - instead of 100% equity, it is 100% bonds.
Of course, this needs to be taken with a large pinch of salt. Understanding the long-run nature of human capital is difficult. In addition, few (if any) investors design their glidepaths based purely on the principle of a constant percentage of total capital invested in equities. The assumptions underpinning this result can be challenged and behavioural factors, such as loss aversion, also matter.
However, we believe it does give pause for thought. The basic fact remains that if human capital is substantially equity-like when young,2 it may not make sense for financial capital to be heavily invested in equity. While a dog's destination is tied to his owner, DC investors can take steps to ensure their ultimate outcomes are not overly tied to equity returns.3
This post is funded by LGIM
2. Note our result that younger investors' human capital is more equity-like isn't because younger investors are less settled into their careers or more likely to lose their jobs in a downturn (although both may also be true), but rather because their future contributions extend much further into the future.
3. There is a related issue that the equity risk premium is uncertain. Rob Arnott described the act of concentrating the majority of one's investment portfolio in one investment category, based on an unknowable and fickle long-term equity premium, as a dangerous game of "probability chicken."