Imagine you’re the head of a business unit, one of 25 within a large company. You’re offered the chance to invest in a project with a 50% chance to make £2m, and a 50% chance to lose £1m. Do you take it?
Probably not, right? A £2m win could lead to a nice bonus, but a £1m loss could ruin your career. To your personal framing (shared with your 24 fellow business-unit heads) the gamble is irrational.
Now imagine you’re the CEO of the whole company. Do you want each of your business-unit heads to take the risk? From this portfolio-wide perspective, it’d be irrational not to. The aggregate expected value of the 25 independent bets is overwhelmingly positive, with minimal chance (~5%) of the company losing money overall.
Richard Thaler used this example, building on the work of Matthew Rabin, to show how decisions that seem rational in isolation can be deeply flawed when viewed in a wider (and quite possibly more relevant) context. The same danger arises when we divide investment portfolios into pots, and make decisions about each independently.
In a previous article we took a broad look at the use of pots in portfolio management. We saw:
- That many errors stem from thinking ‘risk tolerance’ is a feature of a portfolio, rather than a person;
- How matching goals to separate pots originated from selling product-specific solutions to isolated problems, something financial planning has largely, and valuably, evolved away from;
- How the benefits of a pot-based approach (motivation, comfort, and clarity) could be replicated virtually, without the accompanying costs;
- The five major downsides of a pot-based approach concealed beneath the surface-level temptation; and
- The additional advantages of a more holistic approach, focusing on financial and psychological flexibility.
Because we encounter it so often in the wild, this article takes a closer look at perhaps the most serious downside: how potting can suppress long-term portfolio growth by encouraging investors to take too little risk.
There are four main reasons why pots, as they are typically used, lead clients to take less risk than is suitable – and consequently earn lower long-term returns than they otherwise might.
1. Underappreciation of the wider financial picture – In line with the CEO example above, optimising for pots means missing out on optimising for the risk diversification between them. When applying separate risk levels to sub-components of overall wealth, people don’t accurately account for diversification benefits, and therefore under-risk their investible assets as a whole. And it isn’t only diversification that gets overlooked. When we focus on each pot in isolation, we also tend to overlook the wider financial resources that contribute to overall Risk Capacity. The result can be yet more caution than is warranted by the investor’s overall financial position.
2. Multiplying the temptation to ‘play it safe’ – Holding a cautious slice of cash within a portfolio – a minor deviation from the technically optimal asset allocation – can be a relatively cost-effective means of purchasing some emotional comfort. However, cash buffers aren’t always set as a fixed percentage. Each pot may have its own minimum (if you’re going to have cash at all, might as well make it a meaningful number), or simply get rounded up to something neater. The result can be more cash than would feel necessary if the portfolio were viewed as a whole.
But cash buffers aren’t the only way potting multiplies caution. Investors may be quite happy to implement the intended risk in more cautious pots, but baulk at the risk required in the more adventurous ones. The resulting barbell is apt to be lopsided in practice: cautious pots stay cautious, while adventurous pots become rather less adventurous than intended. The portfolio as a whole can therefore end up taking less risk than planned.
3. The behavioural consequences of narrow framing – Pots heighten the chance of seeing red – with several pots, some can be losing money even when the portfolio as a whole is comfortably in the green. People are naturally prone to pay more attention to these red ‘danger’ signals than they are to the happy green ones. This attentional tilt results in increased anxiety and an increased pull to de-risk investments to avoid seeing more red in future. Worse, the effect can compound: disproportionate attention to under- (or over-) performing pots can prompt unnecessary action, creating yet more mistakes and anxiety.
4. Overweighting individual goals at the expense of all goals – Pots are typically used to match a given amount of money to a given goal. This can increase the chances of achieving that goal, but at the expense of supporting someone’s wider system of goals. When each pot is managed to deliver a particular cash flow at a particular time, today’s aspirations start to look like fixed liabilities. We are then tempted to allocate more to each goal, at the expense of our uncommitted wealth, so that less risk is needed to achieve it. Repeat across goals and the portfolio quietly becomes under-risked, at the expense of future growth and resilience. A shame, given that goals (valuably!) differ in importance, flexibility and conditionality, and will probably change – often significantly – over the years (with new ones emerging along the way!).
But what about emotional liquidity?
A reasonable counterargument, in favour of judicious use of pots, is that they provide emotional comfort (for several reasons, as noted in my previous article). However, as noted there, these benefits can be obtained through changes in how a portfolio is presented, without complicating its structure or incurring the costs of imposing separate pot-level risk targets.
Moreover, even with the help of technology, more complex structuring means more cost: there are simply more overheads associated with maintaining multiple pots, and each additional cost is an additional return foregone.
The essential, the useful, and the rest
Risk Tolerance belongs to the investor, not to any individual pot. The suitable level of risk to take with their investible assets needs to reflect this, alongside their Risk Capacity and Behavioural Capacity, assessed in the context of their overall wealth. Imposing separate risk targets for individual pots – rather than managing them within an overall suitable risk level – introduces inefficiencies which largely lead to unnecessarily lower risk and lower returns. We see under-risking in portfolio construction, because of unseen diversification benefits, under-appreciation of the role of other financial assets, and the multiplication of caution across individual pots. This can then be compounded by over-protecting individual goals, or by attention being drawn to ‘losing’ assets, as immediate loss-aversion usurps long-term strategy.
Some pots are (or can be) essential: for example, restricted-access assets, like those in a pension, can have legitimately separate suitable risk levels. Some pots are (or can be) useful: for example, siloed ‘play pots’, and goal-orientated pots to the extent they increase motivation (though, as above, the potential behavioural benefits can be achieved through changes in how a portfolio is presented, without needing to change how it’s structured). In all other cases, pots should simply not exist.
Greg B Davies, PhD, is head of behavioural finance at Oxford Risk
