fca2019 wrote:The answer is never, the theory is you live off the gains and not the principal.
Spending just dividends when invested in a broad stock index is theory is indefinitely sustainable but risks the dividends (and stock value) halving or more - and staying down for decades. You only have to look back a little over a century ago for such a situation. Spanish flu etc. and stocks in inflation adjusted terms fell considerably as did dividends. That could have left you having to live on half or less of the income of before and little in the way of re-diverting the depleted capital values into safer alternatives (other than also accepting low levels of income).
Looking at recent index linked gilts (inflation bonds), and the real yields across the board (short to long dated) are pretty close to -2.5%, so for someone with £600K who retired at age 57 and had a additional £8K state pension arising from age 67, they might draw 18K/year (3%) and see that last until age 90.
In the above the capital value was reduced by -2.5%/year in reflection of -2.5% real yields on current index linked gilts. Being inflation bonds the effects of inflation are mitigated, they're inflation adjusted values (the £18K income rises with inflation each year).
If instead perhaps 80% was allocated to a index linked gilt ladder such as above, along with a 20% allocation to stocks and stocks did OK, offsetting the -2.5% overhead of index linked gilts (achieved a 0% after inflation collective reward), then the SWR might be increased to 3.5% (£21K/year inflation adjusted income)
Stock heavy allocations have served well for much of time - until they don't, and there are instances of such periods of 'failures'. The US and UK have been historic good case outcomes, for others stocks have been far less successful. Since the very high inflation and interest rates of the 1970's stocks, bonds, house prices etc. have all done very well, in part having gains uplifted by the transition of high to low interest rates (declining interest rates uplifts prices). From very low inflation/interest rates, forward time that historical flotation effect is more inclined to be absent, or worse, see the opposite side of that coin (declining prices as interest rates rise). A primary risk is that of inflation/interest rates that spike, and from such very low present day levels they don't have to spike that high in order that valuations might halve or more, along with dividends also halving or more along with them.
Drawing 4% from a £600K stock heavy could work out fine, perhaps even seeing none of the capital being depleted - maybe even considerably expanded (great for heirs) but what if its dividend and capital values did halve. Drawing 8% from £300K has a low probability of success. Alternatively seeing £24K/year spending halved to £12K/year may also not be a viable option.
A difficult decision and when presented with such choices along with indecisiveness a good approach is to equally split between the choices. Worst case perhaps seeing a 25% reduction in income rather than half.
Larry Swedroe takes that further and opts for even less stock exposure but in higher risk/reward stock (30% allocation). In a similar vein I quite like the first of these three asset allocations
https://tinyurl.com/qo8a8c6 (scroll down to the Portfolio Growth chart and click/tick the 'Inflation Adjusted' checkbox). Note the settings for that asset allocation include a 4% SWR Yes US data, so a right tail (good/great) outcome market relative to others, as ever there are no guarantees.
PS In US scale, UK FTSE 250 is small cap, and as UK stocks pay a relatively high dividend compared to US stocks might also be considered as being somewhat comparable to US Small Cap Value. Instead of a 10 year treasury bond fund, a simple high street 5 year fixed income ladder, but with the first run in instant access is a reasonable choice IMO. i.e. buy each of 2, 3, 4 and 5 year bond funds, hold each to maturity and roll into another 5 year.