My boiler has been faulty this week, which has set me thinking that modern portfolio theory can be applied not just to financial assets, but to mental states.

What I mean is that having to work distracted from the hassle of thinking about the boiler: will the gasman come? Will he fix it? How much will it cost? The upshot was that I was less agitated than I'd otherwise have been.

What happened here was an application of MPT - an idea which with a little thought (that is, more thought than banks ever gave it) - is still one of the most powerful in finance.

Think of everything you have - car, job, friends, family, central heating, whatever - as assets with payoffs which are sometimes positive, sometimes negative (your car breaks down, you row with your wife and so on). If these payoffs are positively correlated, then your happiness will vary hugely: sometimes everything will go well, sometimes eveything badly. If, however, they are negatively correlated then your mental state will be more stable. And this is what happened to me. My job has a (lowish!) payoff in terms of my well-being. But that payoff is uncorrelated with other negative payoffs in my life, with the result that it stops me getting too down. In effect, my mental states are diversified.

I don't think I'm eccentric here. When men half-joke that they go to work to get some peace from their families, they're expressing a similar idea: the psychological payoff to work is uncorrelated with the psychological payoff to family life, causing one to diversify the other.

This analogy helps illuminate several facts about happiness:

1. Why are the unemployed unhappy, even controlling for income? One reason is that they lose the diversifying power of work for well-being, with the result that otherwise tolerable niggles - noisy neighbours, a rocky marriage or general down-in-the-dumpsness - hurt more. Another reason is that the loss of a job is associated with other losses - not just of income but of social contacts too. It is not single losses which are disastrous, but correlated ones.

2. Why are people who watch lots of TV (often) unhappier than others? It's because watching TV isn't sufficiently engaging an acitivity to distract us from the stresses of the day. It doesn't generate the "flow" that, say, playing guitar does.

3.Why are religious people often happier than others? A big reason is that religion is a form of insurance, an asset that pays off well in bad states of the world, such as bereavement or unemployment, thus preventing big falls in well-being. Those who reply that religious belief is irrational are like those who claim that a financial asset is over-priced; the statement is only relevant if holders of the asset come to believe it.

There is, however, another analogy between MPT and well-being. Diversification doesn't just reduce downside risk, but upside risk too. The serial shagger or bachelor with a rich internal life might never know a broken heart, but he'll never know the glory of true love either.