funduffer wrote:One point on income per unit is the concept of dividend drag. If, like me, you top up your HYP fairly frequently, then the income per unit measure is always behind the underlying income performance. This is because, when you make a top-up, the number of units increases instantaneously, but the income benefits from the top-up can take up to a year to arrive in your HYP.
Thus a regularly topped-up HYP will always look a bit worse than a a static HYP that is never topped up (all things being equal).
That effect and similar effects (on measures such as XIRR() values) tend to be largest in the early days of a HYP, when the top-ups are largest relative to its existing size. They tend to be small once it's been going for the year required to have built up to actually having a sensible annual income-per-income-unit, unless there is a large capital injection (e.g. from an inheritance or a major Lottery win), and are usually pretty negligible by the time the HYP can reasonably be considered to be 'mature'. But you're right that it does depress the measures a bit - it's just that it's usually a pretty small bit.
If you do want to compensate for those effects, credit dividends when they go ex-dividend rather than when they're paid. Note that if you do this, you will occasionally have to retrospectively undo such a credit, when an already-ex-dividend dividend is cancelled - this is rare in normal circumstances (I've only been affected by two such previous events in the getting on for 20 years I've been HYPing - Railtrack when it went bust and BP in the immediate aftermath of its Gulf of Mexico oil spill in 2010), but there was a bit of a burst of such cancellations as the COVID-19 pandemic really took off in late March and early April this year (I was hit by two of them, affecting British Land and HSBC (*), but know of a number of others such as Persimmon).
Or if you really want to account for them in full pedantic detail, create a fake holding of 'dividend cash entitlements' in your portfolio tracking software. When a dividend goes ex-dividend, add the amount of the dividend to the dividend cash entitlements; when it is paid, remove the amount of the dividend from the dividend cash entitlements and add it to the cash holding; when an already-ex-dividend dividend is cancelled (usually an extremely rare event) remove the amount of the dividend from the dividend cash entitlements without any compensating addition to the cash. That allows you to look at rate-of-return figures with and without dividend drag, by choosing whether to regard the dividend cash entitlements as being 'inside' or 'outside' the portfolio. But the technique does have the cost of needing to enter three transactions for every dividend payment in normal circumstances, which I'd imagine just about everyone will regard as a prohibitive cost!
(*) Note that this is just cancellations after or on the day
the dividend concerned went ex-dividend - I've been hit by plenty of others... :-