You might think I've no complaint. Surely slightly worse than expected payoffs on my premium bonds have been more than offset by the fact that stock markets have done better than expected, so overall I'm better off than I might reasonably have expected.
Such a reply is true, but it misses the point. I think of premium bonds and shares in different ways. I regard premium bond prizes as spending money, but I save profits on my equity funds. For me, shares and premium bonds are in different mental accounts. Not getting extra spending money is therefore an inconvenience.
From the point of view of elementary economics textbooks, I'm being irrational. A pound is a pound regardless of where it comes from. These textbooks tell me I should maximise my lifetime utility subject to my wealth constraint, which entails regarding premium bond prizes and stock market profits as the same.
Such advice, however, might be wrong. There's a difference between premium bond prizes and stock market gains. A premium bond prize is permanent wealth: a £50 win doesn't affect my chances of a future prize. But a rising stock market might not be so permanent. Rising share prices might be due to investors believing the market has become less risky and so they require lower expected returns. Worse still they might be due to irrational exuberance in which case prices will fall. If equity gains aren't long lasting, it would be imprudent to spend them.
What's more, it's computationally impossible to maximise utility. And attempts to do so risk being undermined by countless errors of judgment. It would be easy for me to err on the side of overestimating future equity returns - simply because even past ones can only be estimated roughly - and so spend too much now.
To save myself the time and trouble of maximising utility, which is potentially counterproductive, I use simple rules of thumb. There's a mental account into which I put savings: some of my salary plus all returns on my wealth bar premium bonds. And I spend the rest.
Millions of people do much the same. We have a mental jar for savings, another for emergencies, another for fun spending. Some even have actual jars.
And so we should. Mental accounting, far better than textbook utility maximising, saves us thinking time and imposes self-control.
However, as we approach retirement we should change the way we do our mental accounting. While we're in work, most of us allocate money to different jars. But, says Diane Garnick of US pension fund TIAA, when we think about retirement we should fill the jars sequentially. We should first fill up the jar for necessities, then the one for emergencies, and so on.
Say you need £10,000 a year for necessities such as food, utilities and so on. If we assume this spending will come out of taxable income, you'll need around £12,500 a year. You should then ask: how much of a pension pot do I need for this jar? There's an old rule of thumb that says we can spend 4 per cent of our wealth each year in retirement while leaving capital intact. This rule says we should simply multiply that £12,500 by 25, which means we need a pot of £312,500. In an era of low returns, however, this might be too optimistic. If we replace the 4 per cent rule with a 3 per cent rule, we must multiply that £12,500 by 33, implying we need £412,500.
We'll need an emergency fund to pay for car or home repairs or help out indigent relatives. If we call this £5,000 a year pre-tax we need another pot of £165,000.
And then of course we need to spend money on having fun. If we call this £12,500 pre-tax a year we need another £412,500.
Adding these up gives us the need for a pot of almost £1m. But on the other hand, the state pension gives us just over £6,000 a year, so our rule of 33 means we can knock off £200,000. And we can knock off more if we have some final-salary pensions. And we can knock off even more if we plan to run down our wealth and not leave a bequest.
We should do this sort of calculation before retiring. As Ms Garnick says, "Waiting until retirement to think through spending strategies is akin to planning for parenthood while you're in labour."
And we should guard against overoptimism. It's easy to persuade ourselves that we'll live frugally in old age and so can get by on a little. There's a germ of truth in this: we'll save some work-related expenses such as transport costs. But, on the other hand, we might find ourselves spending more when we have more time on our hands.
We must also remember than in retirement we have less margin for error. If you spend too much while you're working you can make up for it by working longer and earning more. In retirement this is less easy. Sure, you could stack shelves at B&Q - and increasing numbers of older people do. For me, though, having to do as you're told defeats the very object of retiring. This argues for erring on the cautious side.
Equally, though, we should remember a big fact: retired people are generally happier than middle-aged ones. This tells us that recent retirees haven't had to rein in their spending so much as to cause misery, which implies that people generally haven't substantially underestimated their retirement needs. So why should they suddenly start to do so?
It's possible to save too much, and to be irrationally cautious as well as too reckless. It's probably impossible to get the balance right. But perhaps mental accounting can help us do so more than textbook fantasies of utility maximising.
