The traditional 'asset allocation' industry typically makes 'investor risk appetite' your problem, not theirs. They then perform this outdated pre-MPT analysis of the kinds of asset your risk profile might need. In reality, you need them all, in toto, and your risk appetite only drives the degree of leverage on that total portfolio. Secondly, using some nineteenth century maths on annuities and perpetuities, they take your requirement of needing a fixed amount at a date in the future, run the formula, and work out what your monthly premium ought to be to achieve that future cashflow. Note this too works only by eliminating all asset types and strategies except relatively safe loans/bonds, together with a hope that inflation doesn't destroy the real future value. However, if you're willing to accept uncertainty in the primary return stream (which becomes increasingly OK the longer your relevant time horizon is), then you can replace a safe (close to risk free) return with increasingly risky returns.
But I think one should try to build a model of the risk appetite, which is to say a model of the wealth process. This would be a rather complex process. Stochastic no doubt, and with feedback from the actual experienced output of your core investment model. It is much grander (much more destined to failure too) than knocking off a perpetuity to pay for your children's university bill.
Before doing that, it might be worth thinking if there are any macro or qualitative insights which might be gleaned by thinking about a world where everyone, rich and poor, operated a wealth process. Are there implicit biases in the behaviour of investors based on how wealthy they are? Secondly, how distributed is wealth? How does that matter?
No comments:
Post a Comment