My original aim was to try to identify what might be regarded as the "locked-in" return,  if there were such a thing. However, I have now moved on, to explore an area which might be more interesting to trustees, especially given the likely impact of FRS17.

What I suggest they would like is a scientific estimate of what the ultimate MVR might be after,  say, 10  years (or 15 years). This may sound like a pipe dream (snake oil, even) but DVR appears to have been much more effective in tracking that end-figure than MVR itself.

It is not spot estimates which matter but, rather, the convergence (or divergence) over time.  This suggests that we should use cumulative  statistics. This could be done by targeting the expected fund at the end of the period. However, I have opted to look at annualised returns over the period (always from 31 December until 31 December). 

Most of the charts reached below (for 10 years or 15 years) will show three curves, being:

1. cumulative MVR (% pa) at the end of  the period (straight)
2. cumulative MVR (% pa) over the 10 years
3. cumulative DVR (% pa) over the 10 years

An interesting statistic is the ratio between:

• the area between (1) {white} and (3) {dark blue}   to
• the area between (1) {white} and (2) {light blue}.

Figures lower than 1 imply that DVR tracks the end-result for MVR better than MVR (and  vice-versa). The progress of this statistic over time is shown in a separate chart. It will be seen that DVR normally,  but  not always, tracks the end-MVR better than MVR.

Clear  exceptions will be observed from these  charts. Rather than  “efficiency” (the term originally used), I now refer to “DVR divergence”.

Each chart is based upon UK equities alone OR UK gilts alone. As most UK balanced portfolio are dominated by equities, I have not felt it necessary to show the results for balanced funds but I could make them available. In  all cases,  the  underlying market performance follows the reinvested UK indices without any net new money. While the figures generally appear reasonable for equities, the agreement for gilts is rather more variable for any period length than for equities.

• Bonds are assumed to be redeemable at par after 15 years.
• Equities are assumed to be sold after 15 years, with the sale price and dividend growth rates both being taken as one-half of the  assessed DVR.
• Cash is taken at face value.

Results for 10 years

Results for 15 years