I should have said that doesn't exclude using cost averaging in addition to AIM. If for instance you had $1M already and won another $4M from the lottery (lump sum), then a reasonable choice might be to cost average into the market over 3 years so as to reduce the risk of having bought all-in at what later transpired to be a relatively high price. Cost averaging a total $5M into AIM-HI (80/20 stock/cash) = $1M cash, $4M stock, which over time AIM will further cost-average via trading. Such combined cost-averaging is more likely to result in having 'bought' at an overall 'average' price and as such is more likely to provide average (or better) rewards.
Someone else who had $1M, won another $4M and dumped the lot into the market (buy and hold) at day zero might later see that they bought all-in at a peak, and might not even see inflation pacing rewards (depending on end-date chosen).
In all but the more exceptional straight line upwards (strong Bull) cases, 80/20 periodically rebalanced stock/bonds will tend to compare quite closely to 100% stock rewards - ON AVERAGE. For some subset periods however 80/20 will be better than 100% stock.