1. Empirical research. I did look into this and a quick google scholar search reveals lots of sceptical papers.

2. The theory doesn’t make sense. On the efficient markets view, at any time that you put money into the stock market, you would expect the market to rise, just because the stock market increases over time in expectation. If you have £1m, then you would always be better off in expectation putting that into the stock market as a lump sum. If you spaced it out as 100 £10k investments over 100 months, then, in expectation you would miss out on the investment returns you could have got in those 100 months.

Interesting, thanks. For point 2 - is there some trade-off from the instalments being spread over time against the time they’re in the market for? On one hand, you have investing £10K in one payment and leaving it for 5 years, and on the other hand investing £2k each year for 5 years. But could there be a middle-ground, some function of the variability and average returns, e.g. spreading the £10k in monthly payments in the first year, then leaving it for the next four, that does better than both extremes?

For point 1 - from very quickly skimming these papers (and as an amateur) it looks like the pound-cost averaging approach is beaten by other more complex approaches, but it still seems to be better than lump-sum. Is that your understanding?

I’m very sceptical of pound-cost averaging.

1. Empirical research. I did look into this and a quick google scholar search reveals lots of sceptical papers.

2. The theory doesn’t make sense. On the efficient markets view, at any time that you put money into the stock market, you would expect the market to rise, just because the stock market increases over time in expectation. If you have £1m, then you would always be better off in expectation putting that into the stock market as a lump sum. If you spaced it out as 100 £10k investments over 100 months, then, in expectation you would miss out on the investment returns you could have got in those 100 months.

Interesting, thanks. For point 2 - is there some trade-off from the instalments being spread over time against the time they’re in the market for? On one hand, you have investing £10K in one payment and leaving it for 5 years, and on the other hand investing £2k each year for 5 years. But could there be a middle-ground, some function of the variability and average returns, e.g. spreading the £10k in monthly payments in the first year, then leaving it for the next four, that does better than both extremes?

For point 1 - from very quickly skimming these papers (and as an amateur) it looks like the pound-cost averaging approach is beaten by other more complex approaches, but it still seems to be better than lump-sum. Is that your understanding?

Is there a timing approach you’d recommend?

Someone shared this link with me which supports your view that lump-sum is generally better, especially if you don’t have diminishing utility.