It’s a good rule in life to learn from the mistakes of others if you can. So what could the board of a brand new pension fund learn from others to make sure its internal design and external relationships are as robust as possible?

Like a vast social experiment, providing pensions for future older generations is an exercise fraught with uncertainty. Outside the financial world, who would care to bet on mortality rates far into the future, let alone interest rates, inflation and equity market returns? While they might not be called on to make accurate forecasts of these factors in years hence, pension fund boards do make these judgements, or at least approve them, when they set asset allocation strategy and make decisions about liability hedging. Is our imaginary new pension fund checking that it has the right people on board to make these primary assumptions and that it is relying on sound advice?

An important lesson would be to make the trustees of pension funds aware of the biases in their decision making. Theo Kocken of the Dutch risk and fiduciary manager Cardano calls it aversion to a sure loss. Others might call it optimism. Whatever behavioural finance epithet you like to attribute to it, many of us find it hard to know whether and when to cut our losses.

A December 2011 paper from the CFA Society of the UK entitled ‘Financial Amnesia’ pointed out that behavioural traits, combined with other factors like defects in the principal-agent relationship and moral hazard, play an important role in financial failures. One is a failure to dismiss evidence that refutes a valued notion – cognitive dissonance. In our case that could be challenging the assumption that the equity risk premium will bail the pension fund out.

Another is the illusion of control. AIG’s chief risk officer dismissed the risk inherent in his company’s half a trillion dollars of notional CDS exposure. Likewise, Orange County’s treasurer was hailed as a financial genius until the Fed put paid to his complex investment strategy by raising interest rates.

There is currently a debate on who is best equipped to provide pensions. In the Netherlands, the regulator has, for some years, stated that the best deliverers of pension promises are large entities. The UK’s pension regulator stated in October that it would issue a code of best practice for employers to choose DC auto-enrolment.

It’s worth noting what goes wrong when you don’t get these fundamental elements of the pension deal right. The UK’s National Employment Savings Trust (NEST) has deliberately avoided putting the word ‘pension’ in its name, so toxic is the perceived legacy of mis-sold individual pensions in the 1980s and 90s.

Asset managers are criticised for pushing unsuitable products on individual savers and institutional investors alike. But part of the problem has been an over-reliance on equity risk that became embedded in the culture of UK pensions, both occupational and retail. One of the reasons why DC contributions are so low in the UK is because sponsors were advised that high equity returns would make up the difference.

All the while, DB pensioners in the UK public and private sector enjoy gold-plated inflation-proof pensions. Their children, in fact most people born after the 1960s or 70s, will enjoy a much less certain future than they did. In the Netherlands, a painful rights-cutting exercise is accepted by most as a necessary evil to stabilise the system for future generations. So an important lesson would be to ensure some flexibility in the pension contract.

But perhaps the most important lesson that our imaginary pension fund could draw would be to know its own shortcomings – in the promises it makes, and in its resources, both financial and human.